451 results about "Time series modeling" patented technology

The invention discloses a water quality time sequence prediction method of an SSA-VMD-LSTM-XGBoost hybrid model, and belongs to the technical field of water quality monitoring and prediction. Comprising the following steps: (1) data preparation and preprocessing; (2) optimizing the water quality time sequence decomposition of the VMD based on SSA: optimizing a penalty factor and a modal number of the VMD by adopting a sparrow search algorithm (SSA), and decomposing the water quality time sequence into a plurality of sub-components with high stability and low complexity by utilizing the optimized VMD; (3) construction and training of an LSTM-XGBoost hybrid prediction model: constructing a hybrid prediction model fusing long-short term memory (LSTM) and extreme gradient boost (XGBoost), inputting a high-frequency component into the LSTM model, inputting a low-frequency component into the XGBoost model, and finally performing superposition and integration on prediction results of the models; and (4) multi-component prediction result integration and performance verification. According to the method, adaptive optimization of VMD parameters is realized through SSA, the feature extraction and time sequence modeling capability is improved by combining the advantages of LSTM and XGBoost, and the prediction precision and stability of the water quality time sequence are effectively improved.

The invention relates to the technical field of robot control, and discloses a master-slave cooperation-based control method, device, equipment and medium, and the method comprises the steps: collecting a spatial pose, a contact force and a dual-view image, and constructing a multi-modal demonstration data set; executing time alignment and normalization to form a standardized input sequence; action features, force features and visual features are extracted through the action modal coding network, the mechanical modal coding network and the visual coding network; fusing to form a multi-modal tensor sequence, and inputting the multi-modal tensor sequence into a Transform decoder to generate a joint time sequence representation; outputting a training action prediction vector from the joint time sequence representation through an action predictor, constructing a supervision loss function in combination with expert demonstration annotation to update each network, and obtaining an optimization control model; and processing the real-time input control driving execution mechanism based on the optimization control model. According to the method, the control model is trained and optimized through multi-modal sensing and time sequence modeling, the joint control instruction is generated in deployment, and stable execution of a complex interaction task is achieved.

The invention discloses a cloud data center full life cycle monitoring and early warning system based on digital twinning, and the system comprises a multi-source collection and preprocessing module which is used for collecting and preprocessing multi-source data; the twinborn modeling and simulation module is used for constructing a digital twinborn body and completing virtual-real mapping and mechanism simulation; the time sequence modeling module is used for constructing a state space-Kalman enhancement time sequence evolution twinborn model and carrying out continuous time sequence modeling; the virtual-real residual error self-correction closed loop module is used for generating a virtual-real residual error and updating the digital twinborn body and time sequence evolution twinborn model; and the risk and linkage module is used for executing risk identification, grading and life cycle linkage according to the corrected time sequence characteristic representation. According to the invention, a state space-Kalman enhanced time sequence evolution twinborn method is adopted, virtual-real self-correction and full-life-cycle early warning of the cloud data center are realized, and the method has the advantages of high precision, strong robustness and self-optimization.

The invention discloses a digital human construction method and device based on a heterogeneous emotion semantic graph and long-sequence emotion modeling, and the method comprises the steps: obtaining multi-modal emotion input data of a text, voice and a visual image, extracting features, and constructing a multi-modal emotion feature set with a timestamp; constructing a heterogeneous emotion semantic graph which comprises user entity nodes, modal feature nodes and emotion concept nodes, modeling a semantic association, state transition and conflict suppression relationship through a multi-type edge structure, and introducing a dynamic evolution and conflict discrimination mechanism; performing time sequence modeling on the emotional state sequence by utilizing a local-global double-layer emotional modeling mechanism, and respectively capturing short-time fluctuation and long-time trend; performing cross-modal fusion on the emotional state and the modal features, and decoding the emotional state and the modal features into behavior parameters for controlling expressions, voices and actions of the digital human; and multi-modal emotion expression of the digital human is driven. Compared with the prior art, the emotion recognition accuracy and expression continuity and naturalness can be effectively improved.

The invention discloses a power grid equipment fault diagnosis method based on time sequence modeling and knowledge enhancement reasoning, and the method comprises the steps: extracting multi-modal features of multi-source data of equipment, and carrying out the fusion of the multi-modal features to obtain a fusion feature vector; on the basis of the fusion feature vector, combining static structured parameters of the equipment, and carrying out combined modeling to obtain a state evolution coding vector; based on the fusion feature vector and the state evolution coding vector, multi-path reasoning based on knowledge enhancement is adopted, and the fault confidence degree of each reasoning path is obtained; and carrying out interpretable fusion and adaptive optimization on each fault confidence coefficient to obtain an optimal fault diagnosis strategy. According to the method, the whole process of running state evolution is accurately depicted through fusion of multi-modal features, structuring and time sequence joint modeling; knowledge-enhanced multi-path reasoning is adopted, effective integration of data driving and expert experience knowledge is realized, interpretable fusion and adaptive optimization are carried out, and the accuracy and reliability of fault diagnosis are greatly improved.

The invention discloses a verifiable APT attack chain extraction method based on a cognitive distillation and gating technology. The method comprises the steps of input processing and extraction, dynamic gating weight generation, hierarchical prototype decision making, adversarial robustness guarantee and output. An attack chain time sequence modeling and dynamic gating mechanism is created for the first time, causal constraints are injected through phase perception position coding, semantic fragmentation is avoided through full-word masks, and splitting is conducted in the cracking stage; historical dependence and semantic similarity weight reduction are fused to block noise, and the cutting precision is improved. A cognitive distillation dynamic knowledge evolution system is constructed based on prompt engineering prototype matching and self-attention feature decoupling double engines, so that few-sample reasoning and decision determinacy quantification are realized, technical representation is stripped, a kernel is focused and attacked, and real-time evolution of a knowledge base is promoted; a defense closed loop is constructed by parameterized noise injection and KL divergence, a verifiable countermeasure immune system is established, an attack is simulated to optimize a defense benchmark, probability distribution stability is monitored, the misjudgment rate is extremely low, and a traceable audit log is output.

The invention discloses a probiotic packaging stability control method based on deep learning, and the method comprises the following steps: collecting multi-source data in a packaging process, and carrying out the preprocessing of the multi-source data; performing time sequence feature analysis on the standardized structured data set, and extracting multi-scale time sequence features; performing target segmentation and multi-dimensional structure feature extraction on the standardized microscopic image data set; feature fusion is carried out, a probiotic packaging stability feature space is constructed, and a stable feature vector is generated; inputting a time sequence prediction network, and carrying out time sequence modeling and prediction processing; performing control parameter adjustment and constraint optimization based on a difference value between a prediction result and stability reference data; and executing and updating the packaging control adjustment scheme according to feedback to complete a packaging control closed loop. According to the method, deep learning and multi-source data analysis are fused, intelligent prediction and adaptive control of the probiotic packaging process are realized, and the method has the advantages of high stability and high precision.

The invention discloses a shared unmanned aerial vehicle task scheduling method based on big data analysis, and the method comprises the following steps: collecting multi-source data of a shared unmanned aerial vehicle platform, and carrying out the preprocessing; performing time sequence feature extraction and multi-dimensional feature fusion processing; inputting an improved Crossform model, sequentially carrying out feature coding, time sequence modeling and multi-task prediction processing, and then carrying out feature splicing; task request information is extracted, and feature fusion, correlation analysis and matching score calculation are carried out in combination with a task energy consumption comprehensive result; establishing a multi-objective optimization function to perform state and action modeling and reinforcement learning iterative optimization processing; a task instruction is generated and issued, feedback is collected and executed, an optimal scheduling strategy is updated in real time, and the unmanned aerial vehicle is driven to execute adaptive scheduling. According to the invention, big data analysis and reinforcement learning technologies are fused, intelligent cooperative scheduling of tasks and energy consumption of the shared unmanned aerial vehicle is realized, and the method has the advantages of high efficiency, energy saving and adaptive optimization.

The invention discloses a video crowd counting method based on a cascaded cross-domain feature interaction network. The method comprises the following steps: carrying out data enhancement processing of random cutting and horizontal flipping on a current frame and front and back frames of the current frame; and constructing a cross-domain feature interaction network composed of a spatial domain branch and a frequency domain branch. The frequency domain branch extracts frequency domain feature output of different stages through a high and low frequency signal aggregation module and a feature encoder based on adjacent frames; the spatial domain branch is based on a single-frame image, and static spatial semantic features are extracted through a feature encoder. Cascade fusion is carried out on the double-branch features on multiple scales, two-way channel cross attention is utilized to reconstruct time sequence correlation frequency domain features of a current frame, and fusion and reconstruction of the two domain features are achieved through a cross-domain feature mutual modulation module. And after the reconstructed double-branch features are processed by the fusion network, outputting a crowd density map of the current frame by a density regression head. And after training is completed, storing the optimal model for video crowd counting. According to the invention, through cross-domain feature cascade and bidirectional time sequence modeling, the accuracy and robustness of crowd counting in a video scene are effectively improved.

The invention relates to the field of intelligent evaluation, and particularly discloses a robot joint wear evaluation method and system based on a neural network, and the method comprises the steps: synchronously obtaining an original vibration time sequence and a working condition data time sequence of a robot joint, and constructing a feature extraction network with working condition data as a dynamic condition; and pure degradation feature codes only related to the internal wear state are decoupled and extracted. And then, based on the pure degeneration feature coding sequence, a degeneration trend time sequence modeling network is driven to accurately capture a long-term monotonous evolution law of the degeneration trend time sequence modeling network, and a time sequence hidden state representing the real abrasion degree of the joint is generated. And finally, decoding prediction is carried out on the hidden state of the time sequence, accurate evaluation of the remaining service life of the robot joint is achieved, it can be ensured that the real degradation state of the joint is accurately tracked all the time under the working condition of wide changes, and therefore accurate and intelligent health evaluation of the robot joint is achieved.

The invention discloses an all-weather sound source positioning system and method based on multi-sensor fusion, and the method comprises the steps: firstly, obtaining acoustic sensing data, millimeter wave radar data and infrared thermal imaging data, and enabling each data to have a collection timestamp; then, cross-modal time alignment processing is carried out on the multi-source data to map the multi-source data to a unified time reference, and multi-modal fusion features under a unified time axis are obtained; for each time point in the time-aligned multi-modal fusion features, according to the confidence coefficient of the data of each sensor, adaptive weighted fusion is carried out on the data of different modals, and fused common feature representation is generated; and finally, time sequence modeling and joint reasoning are carried out based on the common feature representations of a plurality of continuous time points, and continuous position information of the sound source in the three-dimensional space is regressed. According to the method, high-precision positioning of three-dimensional positions of a plurality of sound sources is realized, so that the accuracy and the stability of sound source positioning are improved in a complex environment and an all-weather condition.

The invention relates to the technical field of voice semantics, can be applied to business scenes of financial science and technology, medical health and the like, and discloses a voice generation method, device and equipment based on pseudo-autoregression modeling and a medium, and the method comprises the steps: obtaining a training sample containing a text sequence, a prompt voice segment and a target semantic token sequence; performing continuous fragment mask training on the text-to-semantic model to obtain a pseudo-autoregression trained text-to-semantic model; generating candidate speech output by using the text-to-semantic model and the initial semantic-to-acoustic model which are subjected to pseudo-autoregression training, and constructing a preference data pair; updating the semantics-to-acoustics model based on the preference data pair to obtain a preference optimized semantics-to-acoustics model; and generating target voice output based on the target text and the target prompt voice. According to the method, the time sequence modeling capability of the model is enhanced through pseudo-autoregression training, and the voice generation quality is directly optimized through the preference data pair, so that the voice alignment precision and the subjective listening feeling performance are improved.

The invention relates to the technical field of belt conveyor auscultation, and discloses a belt conveyor auscultation abnormity management method based on comparative learning and time sequence modeling, comprising the following steps: step 1, acquiring operation sound and equipment parameters, extracting frame features and acquiring a voiceprint baseline; 2, the carrier roller synchronization frequency is calculated, and period and sideband fingerprint prototype vectors are constructed; step 3, inputting the frame feature and the fingerprint prototype vector into a contrast learning network to obtain a contrast embedded vector and a time sequence feature vector; 4, the time sequence modeling network outputs three types of probabilities and classification original output values; step 5, calculating an abnormal score, generating an abnormal event and determining a grade; step 6, calculating a priority score and distributing a work order; and 7, backfilling the disposal information, calculating the total cost and storing the total cost. According to the invention, accurate detection, hierarchical management and operation and maintenance closed-loop processing of abnormal operation of the belt conveyor are realized.

The invention provides a material detection operation behavior identification method based on multi-modal perception, and belongs to the field of artificial intelligence and electric power operation and maintenance detection, and the method comprises the steps: carrying out the feature extraction of the preprocessed visual, motion and voice modal data, and obtaining visual, motion and voice modal features; the visual, action and voice modal features are input into the trained fine-grained behavior segmentation network, the multi-modal input layer is used for carrying out weighted fusion on the visual modal features, the action modal features and the voice modal features to obtain multi-modal global feature representation, and the time sequence modeling layer is used for carrying out frame-by-frame analysis and labeling on the multi-modal global feature representation to obtain the multi-modal behavior segmentation network. The hidden state sequence is output, and the classification output layer is used for classifying the hidden state sequence and recognizing behavior categories; matching and aligning the behavior type with a preset standard operation instruction, and judging whether the operation behavior is abnormal or not; the invention further provides an identification system. And accurate identification and compliance evaluation of the operation process are realized.

The invention discloses a load prediction method, system and device based on the situation of a power distribution network and a storage medium, and the method comprises the steps: collecting historical load data and real-time monitoring data of the power distribution network, and carrying out the ultra-short-term load prediction; calculating a key load index of the power distribution network based on the real-time monitoring data, evaluating the safety state of the key load index, and generating a preliminary safety situation evaluation result; and fusing the preliminary security situation assessment result and the ultra-short-term load prediction result, and predicting the final regional power distribution network security situation. And optimizing operation parameters of the power distribution network according to the final load prediction result. According to the method, the defect that the non-linear characteristic processing capacity of an error correction technology is insufficient is overcome, and the dynamic correlation analysis and time sequence modeling capacity and the multi-dimensional data integration and application capacity of an existing situation assessment technology are improved.

The invention relates to the technical field of energy storage battery simulation, and discloses an energy storage battery module simulation method and system based on machine learning. The method comprises the steps of collecting battery signal circuit data of a BMS system in charging and discharging circulation of an energy storage battery pack, extracting voltage fluctuation characteristics and current ripple characteristics of an energy storage high-voltage box, and constructing a multi-dimensional dynamic parameter matrix reflecting the internal state of a battery according to time sequence alignment; carrying out dimensionality reduction compression on the matrix through an encoder of the generative adversarial network to generate a battery state hidden variable; dynamic evolution modeling is carried out on hidden variables by utilizing a time sequence convolutional network, and hidden variable tracks of three charge and discharge cycles in the future are predicted; and performing reverse decoding to generate a simulation battery signal, and outputting a simulation data flow containing voltage, current and temperature parameters. According to the method, the multi-dimensional features and the time sequence modeling capability are fused, accurate simulation of the running state of the energy storage battery module is achieved, and the method is suitable for diversified application scenes.

The invention discloses a short video content intelligent generation method based on deep learning, and the method comprises the following steps: obtaining short video data and candidate material frame data, and organizing and generating a candidate material frame sequence; inputting the candidate material frame sequence into an improved Slot-VAE model to generate an object-level potential slot representation set; inputting the short video data into a shot-level semantic coding network, and constructing a double-path latent variable structure; establishing an object-level slot evolution module for time sequence modeling; executing joint training, and updating parameters of the model and the network; inputting the target short video data into a shot-level semantic coding network, and generating a target shot-level semantic slot set and a target object-level slot time sequence potential state sequence; and generating a target short video frame sequence in the object-level decoding network and executing post-processing to generate a short video content intelligent generation result. According to the invention, the lens semantic consistency and the object time sequence continuity are improved.

The invention relates to a multi-view three-dimensional human body posture estimation method and system based on double-flow space-view-time sequence modeling. The method comprises the following steps: inputting a multi-view multi-frame image sequence, and performing human body detection and cutting; performing two-dimensional human body posture estimation on the input image at each view angle and each frame, and extracting image features; constructing an image-attitude bimodal alignment expression; sequentially executing sequential modeling of intra-visual space interaction, cross-visual-angle interaction and intra-visual time interaction on the image flow representation and the attitude flow representation in the same layer to obtain attitude flow fusion features; performing regression on the attitude flow fusion features by using a three-dimensional regression head to obtain three-dimensional skeleton coordinates; training the three-dimensional skeleton by adopting supervised learning; and inputting a to-be-estimated image sequence into the trained model and the regression head to obtain a three-dimensional human body posture estimation result. According to the method, the precision, robustness and deployability of three-dimensional attitude reconstruction can be remarkably improved under the condition of not depending on camera calibration parameters and human body priori.

The invention discloses an irrigation method and system based on a crop growth model, and relates to the technical field of farmland irrigation, and the method comprises the steps: training a growth model based on a time sequence neural network through employing the obtained historical meteorological data, the crop growth process data of a corresponding time period, and the moisture state of a soil root zone; calculating the actual evapotranspiration of the crops by combining the actual moisture state of the soil and the moisture stress coefficient predicted by the growth model, and determining the irrigation amount according to the water deficit and the root depth by combining the current moisture state of the root zone with the stage target moisture content. And the irrigation frequency adaptive to the current environmental condition is calculated by using the meteorological prediction data and the actual evapotranspiration. According to the invention, by introducing the crop growth model based on time sequence modeling, correlation learning among meteorological conditions, soil moisture dynamic changes and crop moisture stress states is realized, so that irrigation decisions can reflect periodic physiological needs of crops on the basis of real-time data.

The invention belongs to the technical field of artificial intelligence and multi-modal emotion calculation, and discloses a multi-modal emotion recognition method based on a Mama state space model and cross-modal self-distillation. Through the organic combination of the efficient sequence modeling capability of the Mamba state space model and the knowledge sharing mechanism of cross-modal self-distillation, the advantages of the state space model in the aspects of time sequence modeling and calculation efficiency are fully played, and meanwhile, the limitation of a single model architecture is made up through a cross-modal attention mechanism; the technical bottlenecks of an existing multi-modal emotion recognition method in the aspects of long sequence processing efficiency, cross-modal information fusion and knowledge transfer sufficiency are effectively solved, and an efficient and reliable technical solution is provided for further development and practical application of the multi-modal emotion recognition technology.

The invention discloses a wind power prediction method and system based on DIRMO and a differentiated objective function, and relates to the field of wind power prediction. In order to solve the defects that time sequence modeling and feature processing are difficult to consider at the same time, the prediction precision is reduced, the model robustness is insufficient and the search efficiency is low due to the fact that a single model is adopted for prediction in existing multi-step power prediction, denoising, normalization and feature extraction are carried out on preprocessed power and wind speed data, and multi-scale features are obtained; performing time sequence modeling on the normalized power and wind speed data to generate a future preliminary prediction result; dividing a multi-step prediction task into a plurality of groups, and converting a multi-output problem into a single-output problem for training and prediction; and carrying out automatic optimization on hyper-parameters of the LightGBM models, training each group of LightGBM models by using the optimized hyper-parameters, and carrying out final power prediction based on a GRU preliminary prediction result and multi-scale features. The method is mainly used for predicting the wind power.

The invention discloses a sudden snow disaster risk degree probability prediction method oriented to insufficient historical time sequences. The method comprises the following steps: step 1), constructing a multi-index risk index based on related data such as historical snow disasters; 2) performing time sequence modeling on the violent snow disaster risk index by adopting an autoregressive integral moving average model, and capturing trend and periodic characteristics of the violent snow disaster risk index; 3) discretizing a continuous risk index of modeling production into a plurality of risk levels, and constructing a state transition probability matrix; step 4), introducing space neighborhood influence and carrier exposure constraint, and constructing a Markov space weighted state transition model; 5) coupling CMIP6 climate scene data, and dynamically adjusting future state transition probability, and 6) fusing ARIMA trend prediction and Markov state transition probability, and outputting future multi-period snowstorm disaster risk level probability distribution.According to the method, the accuracy of snowstorm disaster risk prediction and the modeling ability and prediction adaptability of future climate change scenes are improved.

The invention discloses a multi-robot cooperative work method based on a contrast learning role representation reinforcement learning algorithm, and belongs to the field of multi-agent reinforcement learning. Comprising the following steps: each robot obtains observation information of a current environment, and performs time sequence modeling, role modeling, spatial modeling, action selection, award value obtaining and state value generation respectively; in the process, observation information of each robot, global state information of the current environment, a reward value, role representation and the like are stored in an experience playback pool, the steps are repeated, and finally model optimization is completed by using data stored in the experience playback pool. The intelligent agent behavior homogenization problem and the reputation distribution problem in the multi-robot cooperative work process are solved, cooperative contribution between the robots can be effectively quantified, and therefore the robots are stimulated to work cooperatively better in future tasks, the robots preferentially select behaviors beneficial to the whole system, and the overall operation efficiency is improved.

The invention belongs to the technical field of computer vision and action recognition, and discloses a table tennis swing key frame recognition method and system, and the method comprises the steps: carrying out the posture estimation of a table tennis motion video frame by frame, extracting twelve skeleton key points including shoulders, elbows, wrists, hips, knees and ankles, and carrying out the recognition of the key frames; normalization is completed with the shoulder midpoint as a translation reference and the shoulder width as a scaling factor, and a standardized skeleton sequence is obtained; sequentially inputting the sequence into a three-layer one-dimensional convolutional network to extract local features, adding a learnable position code after convolution output, and then inputting a time sequence modeling network formed by six layers of Transformer encoders to capture a long-time dependency relationship; and finally, outputting five types of results of forehand shooting, forehand swing, backhand shooting, backhand swing and non-key frames through a two-layer full-connection network. The system is composed of a data processing module, a feature modeling module and a key frame recognition module, and stable recognition of table tennis swing key frames can be achieved.

The invention discloses a road crack development prediction method and system based on time sequence images and environmental factors, and belongs to the technical field of road maintenance monitoring and intelligent traffic. The method comprises the following steps: acquiring a road crack multi-time sequence image of a predicted road section and multi-source potential energy field environmental variables such as traffic load, environmental climate and pavement materials in a corresponding time period; calculating a comprehensive feature vector based on the image and the environment quantity; constructing a crack evolution potential energy field according to the comprehensive feature vector, and mapping the potential energy field features into a dynamic heterogeneous graph structure; and carrying out adaptive learning and time sequence modeling on the structure through a dynamic heterogeneous graph architecture search model (PE-DHGAS) driven by potential energy, and outputting road crack development prediction information and images. According to the method, the precision and stability of road crack evolution trend prediction can be improved, the prediction result can be used for risk assessment and maintenance decision making, and scientific and prospective support is provided for a road maintenance department.

The invention discloses a dynamic space-time diagram traffic flow prediction method based on federated learning and a related device, belongs to the technical field of traffic flow prediction, is applied to a client under a federated learning framework, and comprises the following steps: the client generates a dynamic adjacency matrix through a forgetting mechanism based on historical and current traffic data, and extracts node space features; mapping the node features to a unified semantic space through a shared semantic mapping function, aggregating to generate a client prototype, and uploading the client prototype to a server; and receiving cross-client aggregation features generated by the server according to the client prototype, updating local node features, and performing time sequence modeling to output a prediction result. According to the method, the prototype representation of the client level is introduced under the federated learning framework, collaborative optimization of explicit modeling and time sequence prediction of cross-client space-time dependence is realized, and the accuracy of traffic flow prediction in a distributed traffic scene is remarkably improved.

The invention provides a multi-agent trajectory prediction method based on space-time causal interaction modeling, and belongs to the field of behavior prediction of a multi-agent system. The problem of low trajectory prediction precision caused by data deviation and false correlation in the existing method is solved. The method comprises the following steps: extracting two-dimensional position coordinates of a lane center line corresponding to driving of each vehicle; calculating the deviation degree between the vehicle position and the center line of the corresponding lane, modeling the spatial dependency relationship between the vehicles, outputting vehicle spatial interaction characteristics, carrying out time sequence modeling, and extracting vehicle time sequence characteristics; feature coding: outputting lane center line features; fusing the time sequence features and the lane center line features to generate vehicle fusion features; constructing a causal discovery network, and outputting a causal probability matrix; based on the vehicle fusion features and the causal probability matrix, space-time causal interaction features between the vehicles are extracted and input into a trajectory decoder, and multiple possible future trajectory sequences of each vehicle are generated. The method is used in the fields of artificial intelligence, robots and intelligent driving.

The invention belongs to the field of energy storage system evaluation, and particularly provides a compressed air energy storage system state evaluation method based on digital twinning, in the method, a multi-physics field coupling digital twinning model is constructed based on a compressed air energy storage system, and a time sequence modeling module based on an LSTM-Transformer hybrid architecture is built in the digital twinning model; based on the calibrated digital twinborn model, state evaluation is carried out on the compressed air energy storage system in combination with a preset evaluation dimension, simulation deduction data of the digital twinborn model and real-time operation data of a physical system are fused in the evaluation process, and an evaluation result is obtained and output; and based on an evaluation result, triggering an uncertainty quantification result based on an LSTM-Transform architecture to carry out graded state early warning. According to the method, the multi-physics field coupling digital twinborn model is constructed and dynamically calibrated, simulation deduction, real-time data evaluation and graded early warning based on uncertainty quantification are fused, and the compressed air energy storage system state evaluation precision, real-time performance and early warning credibility are improved.

The invention relates to the technical field of artificial intelligence, in particular to a risk assessment model construction method based on multi-modal data fusion and a neural network, and the method comprises the following steps: collecting videos, detecting clothing colors, extracting audio frequency domain features, counting text emotion keywords, and constructing a multi-modal risk feature set; the method comprises the following steps: generating risk transmission weight data, mapping the risk transmission weight data to a causal matrix, carrying out back propagation correction to generate risk transmission weight data, judging conflicts according to weight change to generate risk conflict identification information, reconstructing a causal chain, extracting key factors to construct a risk propagation path network, inputting the risk propagation path network into LSTM (Long Short Term Memory) assessment, and carrying out classification to generate a multi-modal fusion risk assessment model. Composite features are constructed through multi-source signals, an adjustable conduction relation is formed in combination with weight changes, self-correction is triggered according to node symbol differences to clear conflict components to highlight key factors, conduction paths are generated according to an intensity descending order, and evolution expression is formed in time sequence modeling. The model is stable, interpretable and practical guiding significance in cultural heritage protection and cultural scene risk management and control.

The invention discloses a network security situation awareness and automatic response decision-making system based on AI, and the system comprises a multi-source data processing module which is used for collecting multi-source data, executing time synchronization, field standardization and feature extraction, and generating a security event vector sequence; the situation modeling module is used for one-dimensional mapping and cross-source combination to form a situation representation vector; the attack relation modeling module is used for constructing a behavior combination structure and generating attack chain stage probability distribution and path contribution degree; the trend prediction module is used for extracting continuous time slices and inputting a time sequence modeling structure to generate a risk trend prediction result; the response strategy generation module is used for generating an optimal response action; and the closed-loop updating module is used for executing actions and updating parameters of each modeling module according to feedback information. According to the method, the multi-source security events are uniformly modeled based on the KAN network, so that collaborative updating of attack relation depiction, risk prediction and response generation is realized.