2487 results about "Data mapping" patented technology

A method for structuring the latent space of an autoencoder is provided. The method includes analyzing natural language descriptions related to input data; creating language-guided libraries that categorize and abstract data features based on the analyzed descriptions; mapping input data into the categorized and abstracted features within the latent space of the autoencoder; and training the autoencoder to minimize reconstruction loss while adhering to the structure imposed by the language-guided libraries.

The disclosed embodiments provide systems and methods for automated mapping of raw data into a data fabric. An innovative approach leveraging Artificial Intelligence (AI)-powered tools and a data fabric to automate the ingestion, transformation, and integration of raw data into a unified model is introduced. By automating the data mapping process, organizations can reduce reliance on manual methods and accelerate their ability to utilize robust insights for exposure management and attack surface reduction. The disclosed solution provides a scalable architecture for unifying cybersecurity signals across cloud and hybrid environments, enabling real-time decision-making and improved organizational resilience against cyber threats

The invention discloses a road and bridge parameter anomaly detection method, and belongs to the technical field of civil engineering. Comprising the following steps: step 1, establishing space-time relevance and a data mapping relation between monitoring points; 2, self-adaptive updating of the judgment rule is achieved so as to adapt to state evolution of a bridge service stage; step 3, performing intelligent attribution analysis on abnormity in the monitoring data; step 4, carrying out joint identification and comprehensive evaluation on the sudden structural damage and the slow degeneration degradation process; and 5, carrying out quantitative analysis on the deviation degree of the bridge health state, and generating a standardized bridge health index. According to the method, the space-time correlation network and the environment-structure mapping function between the monitoring points are established, and a multi-dimensional feature extraction and dynamic weight distribution mechanism is combined, so that environment interference elimination, abnormal attribution intelligent analysis and joint identification of sudden damage and slow degeneration are realized, and finally, a standardized health index is quantified and generated.

The embodiment of the invention discloses an intelligent contract auditing method and device based on a multi-modal large model and a medium, and relates to the technical field of contract auditing, the method comprises the steps that a to-be-audited target contract document is acquired, multi-modal data in the target contract document is analyzed, and the multi-modal data comprises text data, table data and image data; performing feature extraction on the multi-modal data to obtain multi-modal feature data, mapping the multi-modal feature data to a unified dimension space so as to calculate an association weight between each modality through a cross-modal attention mechanism, establishing cross-modal bidirectional link data, and obtaining multi-modal data; the multi-modal feature data comprises any one or more of text semantic features, table relation maps and image visual features; and based on the multi-modal feature data and the cross-modal bidirectional link data, performing single-modal self-consistency verification and cross-modal contradiction detection on the target contract document to generate auditing information of the target contract document.

The invention relates to a remote education data processing system which comprises the following steps: under a remote teaching task, pre-defining a task intention and a data expectation point; a semantic timestamp and a task binding label are printed on each data fragment; mapping the collected confusion data including silence, eye movement drift and prediction into a unified learning semantic vector; a micro-expression + interactive behavior + time sequence decision path ternary modeling mode is introduced, and a potential cognitive intention corresponding to the feature combination is recognized; teaching context information is fused; constructing a cognitive state mapping model; reasoning a current cognitive state label from multi-modal sensing data; searching intervention track VS effect feedback data in a historical database; generating a predicted intervention behavior sequence by using a sequence modeling algorithm; a dynamic combination suggestion chain including light prompt, content reconstruction, personalized practice and tutoring invitation is adopted; and superposing the cognitive state sequences of all students into a group cognitive trajectory map.

The invention discloses a judicial scene-oriented multi-modal data fusion method and system. The method comprises the following steps: 1) collecting multi-modal data in a judicial application scene and converting the multi-modal data into data in a uniform format; 2) mapping the data to a unified semantic space to realize cross-modal alignment; then associating the multi-modal data to obtain the text description of the same case and the corresponding image evidence as the multi-modal features of the corresponding case; 3) constructing a knowledge graph of a judicial application scene based on the multi-modal features; driving multi-modal data fusion based on the knowledge graph and constructing a multi-modal evidence chain of each entity; 4) quantifying the integrity of the knowledge graph and the reliability of the multi-modal evidence chain according to a preset index, marking abnormal nodes in the knowledge graph according to a quantification result, and adjusting the abnormal nodes; 5) generating a structured knowledge graph according to the knowledge graph and creating a dynamic desensitization report; edges in the structured knowledge graph represent relationships between legal entities, and each relationship is bound with a multi-modal evidence chain.

The invention provides a crane remote instruction response delay detection and in-advance compensation method and system, and relates to the technical field of cranes, and the crane remote instruction response delay detection and in-advance compensation method comprises the following steps: adopting an adaptive space-time alignment algorithm to map real-time operation data to a dynamic knowledge graph, and generating a feature vector; inputting the feature vector into a depth map neural network integrated with a causal reasoning mechanism to generate an incidence matrix; a multi-head attention network with a residual structure is adopted to extract time sequence features; constructing a hybrid decision system based on the delay prediction tensor, and outputting an optimal compensation strategy; and establishing a double-closed-loop evolution system with an online learning capability, and dynamically optimizing a prediction and compensation strategy according to a compensation effect. Through the dynamic knowledge graph, causal reasoning, the multi-head attention network and the double-closed-loop evolution system, the remote instruction response delay can be accurately predicted, effective compensation is carried out, and the real-time performance and safety of remote control of the crane are improved.

The invention relates to the technical field of unmanned aerial vehicles, and discloses an unmanned aerial vehicle battery endurance flight capability prediction system, which comprises a multi-dimensional data acquisition module, a feature mapping module, a prediction module, an optimization module and a feedback optimization module, and can be additionally provided with an early warning module. The multi-dimensional data acquisition module acquires battery data and cleans the battery data to generate standardized data; the feature mapping module maps the data to a feature space, and generates a feature sequence cluster containing a multi-dimensional association relationship by using a time sequence segmentation algorithm; the prediction module divides prediction intervals based on a support vector machine algorithm and extracts prediction indexes; the optimization module generates an endurance prediction strategy by predicting and optimizing the network model; and the feedback optimization module performs multi-source data fusion optimization and outputs a prediction instruction. The early warning module can associate the prediction instruction with the battery health degree, output a grading early warning signal and trigger a response mechanism.

The invention belongs to the field of data analysis, and discloses a data analysis method for engineering consultation digital intelligent management, which comprises the following steps: mapping multi-project heterogeneous data to a unified three-dimensional space-time grid through space-time grid division to form a grid mapping data set; detecting data format and attribute conflicts based on neighborhood correlation analysis, and distributing calibration data through dynamic weight; missing values and abnormal fluctuations are processed by using a time sequence autoregression model and a smoothing technology, and a complete data set with continuous time is constructed by combining linear interpolation and multi-dimensional data fusion; key indexes are extracted to construct a digital mirror image containing spatial positions, time sequences and attribute association, real-time data are synchronized, and outlier grid points are corrected; and multi-scene operation conditions are constructed based on the mirror image data, and a dynamic adjustment scheme is realized. According to the method, heterogeneous data are integrated through space-time grids, digital twinning and multi-scene simulation technologies are combined, and full-period management and risk intelligent decision-making of engineering data are achieved.

The invention discloses an intelligent port scheduling and energy efficiency optimization method based on digital twinning. The method comprises the steps that S1, three-dimensional point cloud data and sensor data of a target port are acquired; s2, constructing a port digital twin model comprising a quay crane model, an AGV model and a storage yard model; s3, agents are constructed for a quay crane model, an AGV model and a storage yard model in the port digital twinborn model, and a multi-agent reinforcement learning environment is built; s4, training each agent in a multi-agent reinforcement learning environment through a multi-agent reinforcement learning algorithm; s5, mapping the real-time task data and state data of the port to each trained agent to simulate the scheduling of the target port, and generating a port scheduling strategy; and S6, realizing operation scheduling of the quay crane equipment, the AGV equipment and the storage yard equipment of the target port based on the port scheduling strategy. According to the invention, the operation efficiency and the energy efficiency level of the port can be improved, and powerful technical support is provided for sustainable development of the intelligent port.

The invention provides a dam safety perception fusion association method based on a multi-modal space-time diagram neural network. The method comprises the following steps: dividing a dam into a plurality of structural units, and mapping various data into a three-dimensional coordinate system; a heterogeneous graph structure is defined, and a dynamic adjacency matrix is calculated based on the real-time stress gradient so as to reflect physical connection, mechanical conduction and geological association relationships among nodes; carrying out fusion modeling on multi-source data in the heterogeneous graph structure by utilizing a multi-modal space-time diagram neural network, constructing a causal inference engine based on an output result of the multi-modal space-time diagram neural network, and updating a three-level modeling system through structural equation modeling, anti-factual inference and dynamic weight to obtain the heterogeneous graph structure. According to the method, the dynamic coupling rule among the dam structure, geology and material states is excavated, cross-modal space-time fusion of manual inspection and sensor monitoring data can be realized, the early recognition capability and early warning accuracy of dam potential safety hazards are improved, and the problems of data islands and insufficient relevance in a traditional monitoring method are effectively solved.

The invention relates to the technical field of cold chain management, and discloses a multi-modal AIGC cold chain logistics abnormal event early warning method and system, and the method comprises the steps: collecting the multi-modal data of a cold chain node through an Internet of Things sensing system, carrying out the data mapping through a cross-modal alignment algorithm, generating a multi-modal monitoring feature sequence, and carrying out the early warning of the abnormal event of the cold chain logistics. According to the method, cold chain network evaluation feature distribution is obtained through high-dimensional feature analysis, local and global abnormal features are extracted by using a space-time diagram convolutional network, an abnormal probability information set is generated, the information set is subjected to multi-strategy deduction and diffusion simulation, prediction data are corrected, and finally accurate abnormal event early warning information is generated. According to the method, the abnormal early warning precision and response speed of cold-chain logistics can be improved, the risk and loss are reduced, the spatial topology and time evolution of a cold-chain network are captured, the method is superior to traditional isolated node detection, and the problem that in the prior art, single monitoring data prediction has the limitation of recognition capacity is solved.

The invention relates to the technical field of collaborative optimization, in particular to a geometric parameter collaborative optimization method of a taper hole machining cutter, which comprises the following steps: by constructing a high-fidelity digital twin model, integrating multi-physics field coupling and machine tool dynamic characteristics based on a finite element method, generating a geometric parameter-performance data mapping set and training and calculating an agent model; outputting a Pareto solution set through multi-target global optimization; and constructing a constraint range based on the solution set, and calling a digital twin model to carry out local optimization to obtain an optimal geometric parameter combination. The method comprises a self-correction mechanism: correcting a material constitutive relation and a friction coefficient through experimental data; staged adaptive learning, NSGA-II and DBSCAN clustering are adopted, and the efficiency is optimized along with the method; and a Pareto stability index and transfer learning are introduced, so that the result robustness and the cross-task reusability are improved. According to the method, high-precision and high-efficiency geometric parameter collaborative optimization of the taper hole machining tool can be realized.

The invention discloses a BIM-driven tunnel construction full life cycle management method and system, and relates to the technical field of industrial data processing and analysis, and the method comprises the steps: building a BIM model covering the full life cycle of a tunnel and a data mapping linkage relation; collecting monitoring data, accessing the monitoring data to the BIM model, and performing dynamic linkage evaluation to obtain a construction state result; and constructing a safety management mechanism according to an evaluation result and feeding back field guidance. The technical problem that in tunnel construction full-life-cycle management, a traditional industrial data processing mode cannot achieve full-life-cycle integration and cross-stage linkage of multi-source industrial data in the stages of design, construction and operation and maintenance is solved. The technical effects of full-life-cycle integration and cross-stage linkage processing of the industrial data in tunnel design, construction and operation and maintenance stages, improvement of the comprehensiveness, collaboration and timeliness of industrial data processing, and accurate acquisition of the construction state to optimize on-site safety management are achieved.

The invention discloses a control method and system for remote monitoring of the Internet of Things, and relates to the technical field of intelligent monitoring of the Internet of Things, and the method comprises the steps: inputting an operation data set of Internet of Things equipment into a space-time diagram convolution model, carrying out the local space-time feature extraction of an edge layer, carrying out the cross-equipment cooperation mode analysis of a cloud layer, and generating an abnormal propagation path; the method comprises the following steps: mapping three-dimensional space coordinate parameters in an operation data set of Internet of Things equipment into nodes of a topological structure, mapping interaction data between the equipment into edges of the topological structure, constructing a dynamic knowledge graph, injecting an abnormal propagation path into the dynamic knowledge graph, and updating a fault influence weight between the nodes by applying an improved graph convolution fusion algorithm. Acquiring propagation risk nodes, and performing dynamic sorting and community clustering analysis on the propagation risk nodes. According to the invention, through the improved graph convolution fusion algorithm and the space-time graph convolution model, the capability of identifying the fault behavior in the Internet of Things equipment is enhanced, and the efficiency of edge and cloud collaborative analysis is improved at the same time.

The invention relates to the technical field of application of information technology in constructional engineering, and discloses a method for intelligently extracting engineering quantity application in highway engineering. The method comprises the following steps: standardizing an engineering scale in highway engineering design to form a standardized engineering scale; establishing a preliminary budget project reference template; establishing a connection relationship between the preliminary budget project reference template and the standardized engineering scale, wherein the connection relationship at least comprises a semantic correspondence relationship between project sections and engineering scale fields, a logic matching rule between quota subitems and engineering parameters, and a self-adaptive adjustment mechanism for table format differences; when a new engineering quantity table is imported, automatically identifying and analyzing a data field and a structure of the new engineering quantity table based on a connection relationship, mapping data in the new engineering quantity corresponding table to a corresponding project section and a quota sub-target of a preliminary budget project reference template through semantic analysis and a logic matching rule; and generating a budget file meeting the cost compilation requirement.

The invention provides a root cause positioning method which can be applied to the technical field of artificial intelligence. The root cause positioning method comprises the following steps: acquiring data in a configuration management database, a network topology tool, a monitoring system and a work order system to form a multi-source heterogeneous data set; performing knowledge extraction on the multi-source heterogeneous data set, extracting equipment attributes, network topological relations, fault event entities and timestamps, and storing the equipment attributes, the network topological relations, the fault event entities and the timestamps as structured knowledge; mapping real-time index data in the structured knowledge into dynamic attributes of an entity, and constructing a dynamic knowledge graph; based on a graph neural network and in combination with time sequence features, learning a time sequence dependency relationship and a propagation path between fault events in the dynamic knowledge graph; and outputting a root cause entity, a confidence score and a fault propagation path of the fault event through a causal inference algorithm in combination with the multi-dimensional evidence. The invention further provides a root cause positioning device and equipment, a storage medium and a program product.

The invention discloses a multi-modal monitoring and anomaly detection method, system and device for a belt conveying system and a medium, and belongs to the technical field of anomaly detection.The method comprises the steps that a plurality of multi-modal sensors are arranged on the whole path along the path of the belt conveying system, and original signals in the running process of the belt conveying system are collected; the original signals are preprocessed; performing environment compensation calculation to generate engineering data; mapping the engineering data to the same space coordinate system of the belt conveying system, and completing feature extraction to obtain feature vectors; multi-modal fusion analysis is carried out on the feature vectors, abnormal state identification and fault level classification are realized, and an early warning signal and a risk level are output; and executing a matched response control strategy according to the early warning signal, the risk level and the corresponding space coordinate system position. According to the method, the abnormal state can be dynamically recognized under the complex working condition, differential response control is achieved, and the recognition capability and processing precision of early-stage, multi-source and dynamic anomalies are improved.

The invention discloses a multi-modal metadata alignment fusion method and device, equipment and a storage medium, and the method comprises the steps: carrying out the metadata extraction of structured data, unstructured text data and image data, and generating multi-modal metadata with semantic annotations; establishing a shared semantic embedding space, and mapping the multi-modal metadata to the shared semantic embedding space for coding to obtain a unified spatial vector; determining an alignment candidate pair from the unified spatial vector through similarity calculation, and identifying a semantic relationship of the alignment candidate pair; and performing conflict detection on the aligned candidate pairs and the corresponding semantic relationships, and resolving conflicts based on weight weighting fusion to obtain a unified metadata system. According to the method, improvement and optimization are carried out from multiple aspects of multi-modal data processing, semantic understanding, alignment accuracy, conflict resolution and the like, the defects in the prior art are overcome, and a more accurate and comprehensive multi-modal metadata alignment fusion result can be provided.

The invention discloses an intelligent question and answer reply generation method and device based on a knowledge graph. The method comprises the steps of obtaining knowledge graph data and performing structured processing to construct a node level table and a multi-modal index table; reconstructing a target problem according to the historical session; calling a large language model through a thinking chain to generate a reply, and embedding a structured fact to inhibit illusion; and an active wake-up mechanism is configured to improve interactivity. According to the method, through multi-modal data mapping and a dynamic updating mechanism, reply of image-text audio-video combination is achieved, and the fact accuracy is guaranteed through node association priority matching; the system explicitly guides the model reasoning process, establishes a traceable logic chain, effectively reduces knowledge illusion while improving the answer real-time performance, and has the characteristics of multi-round dialogue continuity and humanized interaction.

The invention relates to a transformer multi-physics field coupling numerical simulation method and device and a storage medium. The method comprises the following steps: performing coupling calculation on a target working condition by using an electromagnetic, heat, flow and force coupling calculation model of a transformer to obtain transformer loss data and temperature rise data; mapping the temperature rise data to a transformer electromagnetic and force coupling calculation model by using a grid node data mapping method to serve as an initial value, and performing coupling calculation on a target working condition to obtain dynamic data of an internal structural member of the transformer; performing fast Fourier transform on the dynamic data in a preset frequency range to obtain dynamic data under each frequency; and mapping the temperature rise data and the dynamic data under each frequency to a transformer sound and force coupling calculation model as initial values, and carrying out coupling calculation on a target working condition to obtain transformer oil tank vibration and external noise data. Compared with the prior art, the method has the advantages that the calculation amount is effectively reduced, and the accuracy of predicting the internal state of the transformer is improved.

The invention discloses a digital twinning system and a construction method thereof, and relates to the technical field of digital twinning, and the construction method comprises the steps: carrying out the classified collection of multi-source heterogeneous data of a digital twinning entity, carrying out the filtering and denoising through a self-adaptive wavelet threshold, mapping the data into a multi-dimensional feature vector, screening key features, and outputting the key feature data; the method comprises the following steps: establishing a physical mechanism layer based on a general physical rule, defining core physical parameters and a constraint equation, initializing a particle swarm to establish a data driving layer, constructing an LSTM time sequence prediction module, executing particle filter state calibration, and selecting a model to configure a communication protocol to establish a virtual-real interaction layer, so as to realize state mapping and instruction feedback of a digital twin entity and a virtual model; and calculating virtual and real state deviation in real time and performing attribution diagnosis, adjusting model parameters or key features according to a deviation source, generating an optimization parameter combination based on a long time sequence prediction result and performing simulation verification in a virtual environment, and controlling entity parameter adjustment through an instruction feedback channel so as to realize predictive optimization.

The invention discloses a live broadcast system based on AI interaction, and the system specifically comprises a data collection module which is used for obtaining the real-time bullet screen data and interactive behavior data of a user in a live broadcast process; the emotion judgment module is used for determining an emotion tendency state judgment result of the user in the live broadcast process through the real-time bullet screen data and the interactive behavior data; the parameter adjusting module is used for dynamically adjusting an emotion expression parameter of the AI anchor according to a live broadcast content theme and commodity recommendation key information on the basis of the emotion tendency state judgment result; and the style migration module is used for inputting the emotion expression parameters into a style migration model, generating dynamic expression image data and mapping the dynamic expression image data into an image model of an AI anchor. According to the method and the device, dynamic matching between AI anchor image generation and user emotion requirements is realized, the image generation capability of the AI anchor is improved, and the emotion interaction effect between the AI anchor and the user is remarkably enhanced.

The invention discloses an intelligent construction site construction risk early warning system and method based on a BIM technology, and relates to the technical field of the Internet of Things. Multi-dimensional data of the field environment, structure, equipment and personnel are collected through a multi-source sensor network and are preprocessed; mapping the data to a BIM model, analyzing a spatial interaction relationship among personnel, machinery and environment by using a space-time diagram neural network, and identifying a periodic risk through an autocorrelation algorithm; calculating an index weight in combination with a dynamic weight distribution algorithm, and constructing a risk value calculation model; risk levels are analyzed and judged according to three conditions that personnel are at a mechanical operation position, under mechanical operation and no personnel are on site; a grading response strategy is adopted according to the risk grade; and optimizing an early warning threshold value through a threshold value judgment algorithm and an LSTM model by using a block chain evidence storage risk event, and iteratively updating a prediction algorithm.

The invention discloses a charging pile dynamic load balancing scheduling system based on reinforcement learning, and the system comprises a data collection module which comprises 5G communication units installed at the charging piles, and is used for obtaining the power of the charging piles, vehicle battery parameters and power grid load data in real time; the reinforcement learning decision module maps the data in the data acquisition module into a state vector through a feature extraction network, and generates a power distribution strategy by adopting a PPO + DQN hybrid algorithm; the communication protocol module realizes state synchronization and control instruction transmission between the charging piles based on an OCPP2.0 standard; the execution control module converts the control instruction into a PWM control signal, and adjusts the output power of each charging pile in real time; through the reinforcement learning decision module, the system can analyze the charging pile power, the vehicle battery parameters and the power grid load data in real time, generate an accurate power distribution strategy, ensure accurate matching of the charging pile power and the electric vehicle demand power, and thus significantly improve the charging efficiency.

The invention is suitable for the technical field of disaster monitoring, and provides a power system-oriented low-orbit satellite multi-source sensing disaster monitoring method, which comprises the following steps of: acquiring heterogeneous observation data and time sequence monitoring data, mapping the acquired data into a space-time diagram structure, and generating a multi-dimensional sensing field of a surrounding environment of power equipment; the method comprises the following steps: analyzing disaster characteristics at a low-orbit satellite end through a lightweight federated learning model based on a multi-dimensional sensing field, generating a semantic instruction which can be executed by a machine, distributing the instruction to an unmanned aerial vehicle cluster and an edge computing node through an inter-satellite link, and triggering a self-adaptive observation strategy; predicting a potential disaster chain reaction based on a pre-trained causal inference model; when a disaster chain triggers a threshold value, hierarchical response is activated autonomously; and a self-adaptive monitoring strategy configuration file is generated according to a disaster response result and is used for task planning of a next monitoring period, so that the early warning precision and response timeliness of disasters such as mountain fire and flood are improved, and meanwhile, the stability of communication and power grid operation in an extreme environment is ensured.

The invention discloses a virtual power plant intelligent regulation and control method and system based on artificial intelligence, and belongs to the technical field of electric power system intelligent regulation and control, and the virtual power plant intelligent regulation and control method based on artificial intelligence comprises the following steps: S1, aggregating equipment side data, desensitizing to generate topological codes, and constructing time scale matrix synchronization; s2, constructing a dynamic model by equipment parameters, and mapping real-time data to output a difference map; s3, adding equipment constraints, building a multi-objective function, optimizing a strategy and performing correlation analysis; s4, a wind and light fluctuation overrun trigger RL strategy and an abnormal switching base line generate a mixed instruction; s5, locally verifying the instruction, and correcting and feeding back parameters if the prediction is out of limit; s6, generating a three-dimensional thermodynamic diagram, and displaying an association report and a historical record by AR; s7, aggregating the data to reconstruct the training set, locally fine-tuning the strategy network and performing incremental updating; the method has the beneficial effects that the regulation and control pain point of the virtual power plant is systematically solved, the operation and maintenance cost is reduced, the new energy consumption capability is improved, and the equipment out-of-limit risk is reduced.

The invention relates to the technical field of bridge bearing capacity evaluation, and discloses an existing continuous beam bridge bearing capacity comprehensive evaluation method which comprises the steps that multi-source data are collected to construct a three-dimensional damage voxel field, and an initial finite element model is generated; simulating a heavy load identification control section, and designing an equivalent static load test scheme according to an equivalent internal force principle; executing a field static load test, collecting elastic response data and constructing an actual measurement response vector; calculating errors and iteratively adjusting parameters in a voxel field constraint range to finish model correction; and generating an operation space-time security envelope based on the correction model, and comparing the state in real time to output a control instruction. A three-dimensional damage voxel field is constructed, voxels are used as a normalized data carrier, discrete apparent and internal damage data such as three-dimensional laser scanning point cloud, ultrasonic velocity and rebound strength are mapped into a continuously distributed space field, and the space variability of the rigidity of an existing bridge material can be represented in a refined mode. And the accuracy of describing the physical current situation of the structure by the initial finite element model is improved.

The invention discloses a multi-modal data fusion method and system based on a large model, and relates to the technical field of data fusion, and the method comprises the steps: receiving multi-modal data, and carrying out the noise layering filtering and time-space alignment; mapping data to a large model hidden space through each modal lightweight encoder, extracting initial features by using a single-modal pre-training model, and converting the initial features through an adapter to generate a hidden space vector; each modal hidden space vector and a task cue word are received, a fusion feature matrix is dynamically aggregated and generated through a self-attention mechanism and a cross attention mechanism of a large model, and semantic integration of multi-modal information is realized; taking the fusion feature matrix as a soft label, and learning a cross-modal semantic mapping capability through knowledge distillation; the trained model is deployed to the edge through dynamic quantization and pruning optimization, an optimized feature matrix is input into a task-customized lightweight head network, and a final task result is output in combination with multi-modal context information. The method can break through the semantic gap between modals, and improves the fusion efficiency.

The invention discloses a spatial registration method for tunnel multi-source heterogeneous data, and the method comprises the steps: constructing a lightweight digital twinborn model based on a design parameter library, a construction log library and a real-time monitoring library; mapping the multi-source sensor data in the real-time monitoring library to a digital twin model space node through a space-time synchronization algorithm; probability distribution model-based abnormal value elimination is performed on the geological exploration data, the while-drilling parameters and the geological sketch, and a discrete smooth interpolation algorithm is adopted to generate a surrounding rock mechanical parameter three-dimensional distribution field; the surrounding rock mechanical parameter three-dimensional distribution field serves as input, and front surrounding rock classification probability distribution is dynamically predicted through a machine learning model optimized through transfer learning; based on the three-dimensional point cloud data of the tunnel face in the digital twin model, extracting geometric parameters of the structural plane and constructing a fracture network model constrained by a normal vector; according to the method, the surrounding rock risk location is accurately positioned based on double-threshold triggering, and multi-scale model seamless fusion is realized through gradient constraint topology fusion.