2714 results about "Computational model" patented technology

The invention relates to the technical field of power equipment, and discloses a digital twin operation monitoring system for power equipment, which comprises a data sensing and acquisition system for acquiring key operation parameters of temperature, current, voltage, partial discharge, vibration and humidity of the power equipment in real time, and performing multi-dimensional data acquisition through a sensor and a data transmission module; the state evaluation and prediction system is used for performing equipment health evaluation and residual life prediction by using a prediction model LSTM based on the collected data, and updating a prediction result in real time; provided is a digital twin modeling system. Through the combination of edge calculation, an LSTM model and a digital twinning technology, the precision and real-time performance of health management of power equipment are improved, data quality is optimized through edge calculation, the LSTM model captures an equipment degradation trend, virtual-real fusion is realized through digital twinning, and accurate monitoring and early warning of the health state of the equipment are ensured, so that intelligent operation and maintenance decisions are optimized, the failure rate is reduced, and the safety of power equipment health management is improved. The equipment life is prolonged.

The invention discloses an industrial equipment fault prediction method based on multi-modal data, and belongs to the technical field of specific calculation models, and the method comprises the steps: carrying out the preprocessing according to the collected multi-modal data of the operation of industrial equipment, so as to unify the format of the multi-modal data, and obtaining the structural data; extracting features of the structured data one by one according to data categories, and obtaining a multi-modal fusion feature through a dynamic fusion mechanism; according to the multi-modal fusion features, a fault prediction classification score is obtained through a deep neural network model to perform fault prediction; and when the drift parameter of the multi-modal data is greater than a preset threshold value, performing incremental training on the deep neural network model through the multi-modal data collected in real time to update parameters of the deep neural network model. Through multi-modal data unified processing, dynamic feature fusion, deep neural network modeling and an online learning mechanism, the problems of insufficient multi-modal data fusion, prediction uncertainty quantization deficiency, poor model adaptability and the like are solved.

The invention relates to an intelligent operation and maintenance question-answering system for cable manufacturing equipment, and belongs to the technical field of computer systems based on specific calculation models. The system comprises an edge data acquisition module, a predictive map construction module, a semantic perception module, a question-oriented reasoning module and a question and answer generation module. According to the system, multi-dimensional real-time data in the operation process of equipment is collected and structurally processed, a process knowledge graph is constructed in combination with industry knowledge, and the causal relationship and reasoning parameters in the graph are dynamically updated according to the data trend. A semantic perception module is used for recognizing the problem intention of a user, a semantic weight vector is formed to guide the reasoning process, and a problem-oriented reasoning module is made to execute joint reasoning on the basis of combining real-time data and a knowledge graph and generate an explanatory conclusion. Finally, operation and maintenance suggestions with high readability are output through a question and answer generation module, and the targets of equipment fault intelligent diagnosis, process optimization and man-machine efficient interaction are achieved.

The invention provides a learning performance evaluation-driven teaching management method constructed based on a capability graph. The method comprises the following steps: S1, constructing a multi-dimensional capability graph; s2, a step of operating a multi-modal data acquisition system; s3, performing a dynamic capability value calculation model; s4, a step of constructing a real-time capability early warning system; s5, generating a self-adaptive teaching strategy; s6, a step of carrying out interdisciplinary ability association analysis; s7, generating a personalized learning path; s8, dynamically evaluating the teaching effect; and S9, dynamically optimizing the teaching resources. According to the method, the core problems of data lag, inaccuracy in intervention and one-sided evaluation in traditional education management are solved by constructing a dynamic capability evaluation model, and a quantifiable decision support system is provided for precise teaching. The innovation of the method is that differential equation modeling and reinforcement learning are combined, and continuous optimization and adaptive adjustment of the education process are realized.

The invention discloses an industrial system automatic fault diagnosis method based on a large language model. According to the method, a three-layer mapping system of industrial data, natural language description and knowledge reasoning is constructed, field multi-source sensor data are subjected to semantic conversion, and a quantitative calculation model based on a large language model is constructed based on historical data and logs. And a fault case is matched in real time with the help of a retrieval-enhancement generation technology to serve as a reference, the fault case and abnormal information are input into a knowledge reasoning model based on a large language model together, a structured logical reasoning chain is generated, and a diagnosis conclusion containing candidate faults, cause analysis and disposal suggestions is further output. Meanwhile, through user feedback and a reinforcement learning mechanism, the model and the knowledge base are adaptively updated, the defects of traditional static rules and expert experience are effectively overcome, the accuracy, interpretability and robustness of fault detection are remarkably improved, and the method adapts to complex and changeable working condition requirements.

The invention relates to the technical field of spraying process control, and discloses a spraying process self-adaptive adjustment method based on temperature measurement. The method comprises the steps that target coating parameters and base material physical property data are obtained, and initial spraying control parameters are generated through a first intelligent calculation model in combination with historical process data and real-time environment monitoring data; when spraying is executed, heat distribution data of a spraying area are collected in real time through a temperature sensing device, and according to the difference between the heat distribution data and a preset target temperature interval, technological parameters are dynamically corrected through a first self-adaptive regulation and control algorithm; meanwhile, a machine vision system is used for capturing actual coating morphological characteristics, and after the actual coating morphological characteristics are compared with target parameters, excitation parameters are adjusted in a partitioned mode through a second self-adaptive regulation and control algorithm so as to optimize uniformity; and collecting whole-process data, comprehensively evaluating the whole-process data through the second intelligent calculation model to generate a process optimization instruction, and updating the first intelligent calculation model and the first self-adaptive regulation and control algorithm parameters according to the process optimization instruction.

The invention relates to the technical field of electric power monitoring analysis, and discloses an intelligent analysis system for electric power monitoring data based on a mutual inductor. The system comprises a mutual inductor data acquisition and structuring module which acquires a current waveform, a voltage waveform and a harmonic component in real time, and generates a standardized monitoring data unit through preprocessing, field mapping and association identification establishment; the theoretical monitoring value calculation module constructs a dynamic calculation model according to historical power data, and an input standard data unit outputs a theoretical value; the rule conformity verification module is used for matching the equipment type with the operation rule and generating single compliance judgment; a multi-dimensional difference analysis module compares a theoretical value with a measured value from time domain deviation, frequency domain deviation and waveform distortion, and generates an equipment-level difference coefficient matrix; the association network construction module is used for constructing a multi-monitoring-point association map according to the equipment identifier and the position information; and the anomaly positioning and strategy generation module combines the map, compliance judgment and a difference matrix, calculates a risk score, generates an anomaly probability distribution map, positions anomaly and matches a monitoring strategy.

The invention discloses a scene topology understanding method and device, a storage medium and a program product, and relates to the field of computer systems based on a specific calculation model, and the method comprises the steps: inputting a multi-view environment image set into a backbone network, and generating bird's-eye view features corresponding to the environment image set; calculating a spatial transformation matrix of the aerial view features at the current moment and the aerial view features of the previous K frames, and performing space-time alignment on the obtained K + 1 frames of aerial view features to obtain multi-frame fusion features; inputting the multi-frame fusion features into a map prior model to obtain aerial view correction features; decoding the aerial view correction features based on a topological decoder, and generating a lane topological graph; comparing the lane topological graph with the annotation data, calculating an error loss function, and optimizing network parameters based on the error loss function; and generating an optimized topological graph, and determining a scene topology result based on the optimized topological graph. By implementing the method, the environment topology understanding capability in a complex scene can be improved, and the generation precision of the lane topological graph is optimized.

The invention relates to the technical field of computer systems based on specific calculation models, and discloses a blood oxygen change monitoring algorithm fused with dynamic Bayesian modeling, which comprises the following steps: quantizing signal uncertainty by calculating entropy of pulse waveform harmonic energy distribution; and on the basis of the entropy value, driving a dynamic Bayesian network to carry out confidence coefficient evaluation and weighting processing on the blood oxygen estimation model, and finally outputting a decision pair containing a blood oxygen estimation value and the confidence coefficient thereof. According to the method, the signal uncertainty is converted into a computable entropy index, so that the system can autonomously distinguish real physiological changes and measurement noise, the problem of misjudgment caused by motion interference in traditional blood oxygen monitoring is avoided, meanwhile, non-inductive personalized calibration is achieved by utilizing the continuous learning ability of the Bayesian network, and the accuracy of the system is improved. And the reliability and the practicability of the medical wearable equipment are remarkably improved.

The invention discloses a distributed time sequence library data management method supporting cold and hot data level-to-level management, and relates to the technical field of computer databases, comprising: receiving a write-in request of time sequence data, and executing preliminary write-in data aggregation and sorting; dynamically identifying data cold and hot attributes based on a multi-dimensional data cold and hot degree calculation model in combination with a write-in behavior and an access behavior of time series data; according to the cold and hot attribute recognition result, in combination with the hierarchy boundary of self-adaptive division, automatic hierarchical storage of the time series data is executed; based on a cold and hot data dynamic migration and scheduling mechanism, according to the access frequency and time decay characteristics of the time series data, dynamically adjusting the storage hierarchy of the time series data, and executing a data migration task; and constructing a hierarchical index system adaptive to different cold and hot attribute time series data, and combining query frequency dynamic identification and index structure automatic upgrading. By adopting a cold and hot data automatic identification and hierarchical storage mechanism, the overall performance and the resource utilization rate of the system are improved.

Systems and methods are disclosed for simulation of occluded arteries and / or optimization of occlusion-based treatments. One method includes obtaining a patient-specific anatomic model of a patient's vasculature; obtaining an initial computational model of blood flow through the patient's vasculature based on the patient-specific anatomic model; obtaining a post-treatment computational model by modifying portions of the initial computational model based on an occlusion-based treatment; generating a pre-treatment blood flow characteristic using the initial computational model or computing a post-treatment blood flow using the post-treatment computational model; and outputting a representation of the pre-treatment blood flow characteristic or the post-treatment blood flow characteristic.

The invention belongs to the technical field of mobile communication, and particularly relates to a task unloading and resource allocation method for edge computing. According to the method, a three-layer network structure is established, a distributed decision framework is constructed through reinforcement learning, the task emergency degree is dynamically evaluated, and computing resources are distributed in a differentiated mode; and task unloading and resource allocation are optimized in combination with an edge-cloud collaborative architecture, so that calculation load balancing is realized. According to the method, aiming at a cloud edge-end collaborative edge calculation model, the total cost of a system is defined as a joint optimization problem of task unloading time delay and energy consumption, the problem model is converted into a Markov decision process, and multi-agent and multi-user oriented deep reinforcement learning algorithm agent near-end strategy optimization (MAPPO) is designed; and obtaining an optimal unloading decision through mutual learning among multiple agents. According to the method, the total cost of the system can be effectively reduced, the rationality of edge computing task unloading and resource allocation decision is realized, and meanwhile, the use experience of a user can be improved.

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 provides a multi-agent system for water quality purification material development, and belongs to the technical field of computer systems based on specific calculation models. The system comprises a data extraction agent, a material design agent, a material evaluation and screening agent, a synthesis method generation agent, a material characterization agent, a working condition matching agent, an effect verification agent, a mechanism mining agent and a material field knowledge base. According to the system, through cooperative work of a plurality of intelligent agents, a development report including structure information, an evaluation report, a synthesis scheme, a characterization scheme, a working condition matching scheme, an effect verification report and a deep mechanism analysis report of a water quality purification material for realizing a target task of the water quality purification material is directly generated according to the target task of the water quality purification material; the method does not depend on manpower and computing power resources, and saves time and labor.

The invention provides a multi-modal knowledge graph completion model training method, a completion method and equipment, and relates to the field of computer systems based on specific calculation models. The method comprises the following steps of: extracting an image block feature vector, a word feature vector, a structure feature vector, a corresponding visual mark, a text mark and a structure mark by adopting a multi-modal knowledge graph completion model; based on each visual mark, the text mark and the structure mark, acquiring multi-modal fusion feature data of each entity by a graph attention mechanism; and complementing the original multi-modal knowledge graph according to each piece of multi-modal fusion feature data, and calculating the target loss of the complemented multi-modal knowledge graph to optimize the multi-modal knowledge graph complementing model. According to the method, fine-grained multi-modal feature extraction and application can be realized in the completion model training process, the reliability and effectiveness of the completion model training process can be effectively improved, and then the accuracy and reliability of multi-modal knowledge graph completion can be improved.

The invention discloses an air flotation vacuum pump performance prediction system based on a specific calculation model, and relates to the technical field of mechanical engineering. Comprising a multi-parameter fluctuation sensing module, a weak feature sensitive extraction module, an abnormal coupling atlas construction module, an instability precursor identification module, a prediction model regulation and control module and a model adaptive optimization module, and carrying out fusion in a preset time window, constructing a multi-parameter synchronous fluctuation sensing model, and outputting a synchronous anomaly index. By introducing multi-parameter synchronous fluctuation perception, weak feature extraction and an abnormal coupling map, accurate identification and early warning of the early instability trend of the air floating vacuum pump are realized, and an intelligent system with real-time perception and closed-loop prediction capabilities is constructed by combining model regulation and control and a self-adaptive optimization mechanism, so that the operation safety and stability are improved.

The invention provides a digital twinning application-oriented electrical equipment electromagnetic heat flow coupling rapid calculation method, and belongs to the technical field of electrical digital data processing.The method comprises the steps that firstly, a three-dimensional model of electrical equipment is acquired and preprocessed, and a full-order electromagnetic heat flow coupling calculation model is established and verified through a temperature rise test; generating an experimental point matrix by using a Latin hypercube sampling method, and constructing a current temperature power density relational data set; performing regional division on the power density field by applying a K-means clustering algorithm, and constructing an electromagnetic response surface model through a radial basis function; establishing a heat flow field order reduction model based on an intrinsic orthogonal decomposition technology, and extracting a dominant mode primary function; bidirectional coupling of an electromagnetic field and a heat flow field reduced-order model is achieved, an improved Lagrange multiplier method and a fixed point iteration method are adopted for processing the nonlinear coupling problem, finally, a software development kit supporting an open platform communication unified architecture protocol is packaged, and the electromagnetic heat flow coupling rapid calculation capacity needed by digital twinning application is achieved.

Workloads for executing a graph neural network (GNN) may be divided among various processing units, such as a central processing unit (CPU) and a neural processing unit (NPU). The NPU may include a data processing unit (DPU) and a digital signal processor (DSP). The CPU may perform precomputation, model optimization, hardware optimization, and compilation. For example, the CPU may precompute a parameter matrix and use the parameter matrix as internal parameters of a GNN. The CPU may also perform node padding, approximation computation, or transfer of DSP operations to DPU to optimize the GNN. The CPU may also perform sparsity data compute and storage, vertical fusion of DSP operations and DPU operations, or data quantization to optimize performance of the NPU. The compiled GNN may be provided to the NPU, and the DPU and DSP may perform the operations in the compiled GNN to produce a prediction of the GNN.

A multi-physics computation method and system for digital twin online simulation, relating to the technical field of physics simulation. A multi-physics coupling simulation computation model of a simulated object is established, and the multi-physics coupling simulation computation model is simplified, and the order of a temperature field simulation model is reduced, thereby greatly reducing the amount of computation, and improving computational efficiency; a low-precision dataset is obtained by means of temperature field analysis, so that the data size required for a model is reduced by means of low-precision data, thus reducing modeling costs; and a basic data-driven model is constructed by means of the low-precision dataset and a sample space corresponding to the low-precision dataset, so that a temperature field distribution result can be rapidly outputted, further improving computational efficiency and reducing computational errors.

The invention relates to the technical field of power mode recognition, in particular to a power equipment fault detection method and system based on a deep learning network, and the method comprises the steps: collecting and preprocessing multi-modal data, such as a power time sequence and an equipment image; monitoring an edge node resource state in real time, and adjusting configuration parameters of the lightweight feature extraction model according to a preset rule; a bidirectional cross-modal attention mechanism is adopted, dynamic weighted fusion is carried out on projection feature vectors of different modals by calculating query, key and value representation and attention weight, and a unified fault representation vector is generated; calculating a fault probability by using the representation vector, and adaptively optimizing parameters of a probability calculation model through online incremental learning; and finally, calculating a task priority, and according to the priority and a real-time resource state, determining whether to upload the data to a superior system for finer analysis. According to the method, the real-time performance, the accuracy and the self-adaptive capability of edge end power equipment fault detection can be effectively improved.

The invention discloses a GNSS real-time deception jamming detection method and system based on deep learning, and belongs to the field of satellite navigation signal processing, and the method comprises the steps: collecting Doppler frequency shift signals of a GNSS receiver in real time, and obtaining an initial training data set; an LSTM prediction model with an incremental learning mechanism is constructed, the LSTM prediction model comprises a double-layer LSTM network and a full connection layer, a gating mechanism is introduced into the double-layer LSTM network, and the initial training data set is adopted to add Gaussian white noise to the prediction model for model training; performing sliding window processing on the initial training data set, and extracting a multi-dimensional feature vector; inputting the trained prediction model for prediction to obtain a model prediction value; calculating a residual error between a model predicted value and a real value, calculating a standard deviation of the residual error after eliminating a preheating period, and dynamically setting a threshold value based on the standard deviation of the residual error; and when the absolute value of the residual exceeds a threshold value and the sampling point processed by the sliding window is located in a preset interference time window, judging that a signal interference event occurs.

The invention relates to an improved NSGA-III-based multi-target geographic site selection optimization method and system, belongs to the technical field of optimization in business management and resource planning, and particularly relates to resource allocation and site selection decision making by using a calculation model. The technology aims to solve the problems of slow scheme convergence, non-uniform solution set distribution and low calculation efficiency when multi-target site selection is carried out under limited resources. According to the scheme, geographic space data is preprocessed, and an initial population representing different site selection schemes is generated; carrying out optimization iteration by adopting an improved NSGA-III algorithm, and calculating a plurality of objective functions such as coverage rate, normalized population density and medical service accessibility; and performing individual selection and population updating by utilizing improved reference point generation and vertical distance measurement. The method is mainly used for improving the efficiency and effect of commercial and municipal decisions such as urban planning, medical resource configuration, energy station layout and logistics network optimization.

The invention provides an automatic perception model method and system for illegal access in a network security isolation area, and belongs to the technical field of computer systems based on specific calculation models.The method comprises the steps that firstly, a network topological graph matrix of the security isolation area is constructed, an equipment asset list is established, and then distributed flow collection nodes are deployed to obtain real-time network data; a deep packet detection technology is used for extracting features to establish an equipment behavior baseline library, a multi-target risk assessment function is used for carrying out risk grade division on equipment, a multi-layer perceptron and a time sequence anomaly detection algorithm are used for identifying abnormal communication, and an equipment fingerprint identification mechanism based on physical layer characteristics is established to verify the legality of the identity of the equipment. A security isolation intelligent sensing network model is utilized to analyze network behaviors, a multi-dimensional abnormal scoring system is constructed to calculate risk scores, a response mechanism based on a rule engine is realized, a federal learning technology can be selectively adopted to optimize the model, and an all-dimensional and multi-level illegal access automatic sensing protection system is formed.

The invention discloses a dynamic noise analysis method and device for an urban elevated road and a medium, and belongs to the technical field of noise monitoring and analysis. The method comprises the following steps: acquiring a three-dimensional geographic information model and real-time traffic flow data of a target elevated road; processing the three-dimensional geographic information model based on a sound wave propagation attenuation model to generate a noise propagation path grid; fusing the real-time traffic flow data with a noise source intensity calculation model to generate a dynamic noise source intensity sequence; calculating a real-time equivalent sound pressure level of each grid unit through a noise superposition algorithm in combination with the dynamic noise source intensity sequence and the noise propagation path grid; performing rolling prediction on the real-time equivalent sound pressure level based on a time window sliding mechanism to generate a noise space-time distribution thermodynamic diagram; and according to the noise space-time distribution thermodynamic diagram, identifying a noise exceeding area, and outputting a dynamic noise evaluation report. According to the method, the technical effect of dynamically predicting and analyzing the noise of the elevated road is achieved.

The invention discloses a new energy access lower harbor area transformer insulation performance and service life evaluation method, which comprises the following steps: firstly, establishing a wind and light storage AC / DC hybrid micro-grid equivalent simulation model, and quantifying a composite action mechanism of new energy fluctuation and harmonic characteristics on dynamic loss and hot-spot temperature rise of a transformer; analyzing a harmonic loss mechanism of the transformer in different operation modes; secondly, constructing a quay crane and field crane operation simulation model, and analyzing the harmonic characteristics of bridge crane equipment and the influence of the harmonic characteristics on the harmonic distortion rate of the power distribution network; then, on the basis of the data, a transformer hot-spot temperature dynamic calculation model is established in combination with bidirectional power flow characteristics; and finally, calculating transient temperature distribution of the transformer through a flow field-temperature field coupling method, and constructing a multi-physical field coupled insulation aging rate model to realize evaluation of the residual life of the transformer. The method solves the problem that the traditional method cannot quantify the bidirectional power flow and harmonic wave composite aging effect.

The invention relates to the technical field of data resource management, in particular to an online evaluation method for data resource value, which comprises the following steps of: constructing a multi-dimensional evaluation index set, and acquiring metadata and market dynamic information of to-be-evaluated data in real time through a distributed interface; carrying out weight distribution on the index set based on a dynamic weighting algorithm, generating a weight coefficient set and constructing a value calculation model; performing standardized preprocessing on the collected original data, inputting the data into the model, and calculating an initial estimated value; and dynamically correcting the initial value estimation value based on the timeliness attenuation factor and the data scarcity factor, and generating a final value estimation value and an estimation report. Through multi-dimensional index construction, a dynamic weighting algorithm, standardized preprocessing and two-factor correction, automatic evaluation of the data value is realized, the problems of single dimension, static weight and neglect of timeliness scarcity of a traditional method are solved, the real-time performance and scientificity are improved, and a quantitative decision basis is provided for data capitalization.

The invention discloses a multi-modal interactive fusion virtual reality emotion computing system, which comprises a multi-modal interactive fusion virtual reality emotion computing device, a multi-modal interactive fusion virtual reality emotion computing device and a multi-modal interactive fusion virtual reality emotion computing device, the virtual reality emotion computing device based on multi-modal interaction fusion further comprises a distributed multi-modal sensing device, a user adaptation interaction module, an interaction feedback module, an emotion extraction module, a multi-modal fusion module and a self-adaptation emotion computing model. The multi-modal fusion module is used for carrying out full-dimensional monitoring and intelligent decision making on a complex scene, the user adaptive interaction module can realize accurate identification and response to user requirements, and the multi-modal fusion module is used for making up for information limitation of a single modal, so that effective interaction and accurate identification can be realized on the whole.

The invention relates to the technical field of logistics park management, and particularly discloses a logistics park intelligent early warning and dynamic evaluation method, and the method comprises the steps: carrying out the time-space alignment of vehicle trajectory data, cargo pressure distribution data and environment data collected by a heterogeneous sensor network, and inputting the data into a dynamic threshold adjustment module to generate a self-adaptive warning threshold; the dynamic threshold adjustment module establishes a dynamic calculation model containing the equipment utilization rate and the vehicle density based on the matching relationship between the historical operation mode and the real-time operation state; based on a multi-dimensional evaluation system constructed based on a self-adaptive warning threshold, generating a comprehensive evaluation index through collaborative analysis of three groups of indexes including transportation efficiency, safety risk and resource utilization; and when the comprehensive evaluation index deviates from the preset range, triggering a grading early warning mechanism associated with the deviation degree. According to the invention, dynamic evaluation and intelligent early warning are realized, and real-time monitoring and intelligent evaluation regulation and control of the operation state of the logistics park are realized through space-time correlation analysis and an adaptive decision-making mechanism of multi-source heterogeneous data.

The invention discloses a target multi-attribute identification method based on feature decoupling and cross-task collaboration, and belongs to the technical field of computers of specific calculation models, and the method comprises the following steps: firstly, extracting the initial features of each task through a lightweight backbone network, carrying out feature decoupling in a subspace, and according to the cross-task feature similarity, carrying out feature extraction; according to the target multi-attribute identification method based on feature decoupling and cross-task collaboration, an orthogonal constraint weight is dynamically adjusted, then mutual information confrontation minimization is adopted to further suppress statistical dependence between tasks, task residual errors are injected in a cross-task feature aggregation stage, differentiation enhancement is achieved, and finally unified joint feature representation is formed. According to the method, subspace statistical independence is realized, independence and necessary collaborative information are considered, a stable basis is provided for subsequent fusion, statistical dependence between tasks is further suppressed through mutual information confrontation minimization, complementation information is reserved while independence is ensured, and feature discrimination and robustness are improved.

The invention provides a deep rock mass creep-seepage coupled near-field dynamics simulation method and system, and relates to the technical field of geotechnical engineering.The method comprises the steps that a near-field dynamics numerical calculation model is established based on geological survey data; based on the Burgers creep constitutive model, a near-field dynamic creep-seepage coupling basic motion equation is formed; and carrying out rock mass creep process simulation considering the seepage effect by utilizing the near-field dynamics numerical calculation model and the creep-seepage coupling basic motion equation, and carrying out iterative solution on stress, strain, displacement, water pressure and damage of each time step in the simulation process until iteration stop conditions are met, so as to obtain the rock mass creep process. And finally, obtaining distribution characteristics of a rock mass displacement field, a damage field and a seepage field and a creep characteristic curve, and realizing near-field dynamic simulation of a rock mass creep-seepage coupling process. According to the method, the technical blank of near-field dynamics in the field is filled, and the long-term stability of the engineering rock mass and the disaster evolution process are accurately reproduced and predicted.