835 results about "Hybrid neural network" patented technology

The invention discloses a smart road driving time prediction method based on multi-source data fusion and deep learning, and the method comprises the steps: constructing a multi-source data collection and fusion platform, developing a spatial-temporal feature deep learning model, deploying a real-time data processing and anomaly detection system, building a self-adaptive prediction optimization mechanism, and constructing a smart traffic application service platform. Comprehensive acquisition, deep analysis, real-time processing, accurate prediction and intelligent service of traffic data are realized. The method specifically comprises the following steps: deploying an Internet of Things sensor network to collect multi-source data, and designing an adaptive weighted fusion algorithm to construct a multi-dimensional dynamic traffic database; constructing an LSTM-GNN hybrid neural network architecture to extract spatio-temporal features; real-time data processing and anomaly detection are carried out by adopting a streaming computing framework and various technologies; designing a multi-objective optimization algorithm and a reinforcement learning model to realize adaptive prediction optimization; and developing a visual decision support system and a personalized navigation service. The system effectively improves the traffic management efficiency and the public travel experience.

The invention belongs to the field of natural language processing crossing, and particularly relates to a visual-text collaborative abstract generation method and system based on multi-modal learning, and the system comprises a multi-modal data receiving module which is used for receiving multi-modal input data in parallel, and the multi-modal input data comprises text data and visual data; the visual semantic understanding module is used for extracting high-level semantic features of the image by adopting a visual semantic understanding model and generating text description; the semantic fusion module is used for carrying out semantic fusion on the original text and the generated image description by adopting a vision-text semantic fusion method based on direction consistency and self-adaptive semantic completion; and the abstract optimization module is used for carrying out semantic reconstruction on the fusion features by adopting a multilayer deep fusion hybrid neural network architecture, and generating an abstract text conforming to context semantics and accurate expression. The method has the advantages that visual and text information is accurately aligned, high-quality abstracts are generated, and the method is particularly suitable for news reports, conference records, video content analysis and other scenes needing cross-modal information fusion.

The invention provides an intelligent power grid relay protection reliability evaluation and lifting control method and system, and relates to the technical field of intelligent power grid relay protection, and the method comprises the steps: collecting the operation data of a relay protection device, carrying out the preprocessing and multi-domain feature extraction, and inputting a deep hybrid neural network to obtain a comprehensive feature vector; a multi-stage fault diagnosis model is utilized to identify a fault mode, a collaborative anomaly and a reliability level at a device level, an interval level and a station level respectively, and a fault propagation path and influence are predicted based on a causal reasoning network. And finally, constructing a hierarchical cooperative control strategy, and through multi-objective optimization, distributed game optimization and hierarchical rolling optimization, realizing coordinated optimization of a protection fixed value and improving the reliability of a relay protection device, thereby reducing the fault propagation probability and ensuring safe and stable operation of a power grid.

The invention relates to a regional pollution process real-time monitoring, regulation and control system and method based on artificial intelligence. The system is composed of a data acquisition module, a communication module, an artificial intelligence analysis module, an intelligent regulation and control module, an execution terminal and a user interaction module. The data acquisition module acquires multi-source data such as pollutant concentration and meteorological parameters by means of a multi-modal sensor, the data are processed and transmitted by the data communication module, and the artificial intelligence analysis module realizes pollution source tracing and accurate prediction by using technologies such as a space-time diagram convolutional network and an LSTM-Transform hybrid neural network. The intelligent regulation and control module generates a regulation and control strategy based on an NSGA-II algorithm, and the execution terminal is responsible for implementation. The user interaction module provides a visual interface and a manual intervention channel. The method comprises the steps of data acquisition and processing, model construction and prediction, regulation and control strategy formulation and execution and feedback optimization closed-loop operation. According to the invention, comprehensive real-time monitoring and accurate regulation and control of regional pollution are realized, the prediction accuracy is improved, the environmental, economic and social benefits are balanced, and an efficient technical means is provided for regional pollution treatment.

The invention belongs to the technical field of bandwidth access management, and particularly relates to an internet access service optimization method and system for adaptive bandwidth management. Through the distributed probe technology, network state parameters and user terminal metadata can be collected in real time, a deep learning network model such as an LSTM-Transformer hybrid neural network is utilized to process and analyze the collected data, the bandwidth requirement in a period of time in the future is predicted, time periodicity and traffic mutation characteristics are considered, and the bandwidth requirement in a period of time in the future is predicted. Equipment association characteristics are introduced, so that the prediction accuracy and reliability are improved, virtual channels are dynamically divided according to the service emergency degree and predicted bandwidth requirements, bandwidth resources are allocated to the virtual channels, it is ensured that high-priority services can obtain enough bandwidth resources in the network peak period, and the network performance is improved. Therefore, the stability and the quality of network service are ensured, and in addition, a bandwidth expansion mechanism and a bandwidth recovery strategy are introduced to cope with the dynamic change of the network bandwidth.

The invention relates to the technical field of equipment fault diagnosis, in particular to an extruder equipment fault recognition method and system based on artificial intelligence, and the method comprises the following steps: collecting key fault features of an extruder in real time based on a multi-mode sensor network, optimizing the signal quality through data preprocessing and feature decoupling, and obtaining a fault recognition result; and the generalization ability of the model is improved by using cross-device feature mapping and transfer learning, a hybrid neural network is combined, a physical constraint layer is embedded on the basis of a data driving layer, a feature incidence matrix conforming to the dynamic characteristics of the extruder is constructed, and a fault prediction model can be adjusted in real time through a dynamic weight distribution mechanism and dual-target loss optimization, so that the fault prediction efficiency is improved. The method adapts to the change of the operation state of the equipment, and realizes the real-time detection, graded early warning and precise operation and maintenance of faults in combination with an intelligent early warning mechanism and a multi-target optimization decision. According to the invention, the operation stability and maintenance efficiency of the extruder equipment are obviously improved, and the method is suitable for equipment health management in the field of intelligent manufacturing.

The invention discloses a medical whole-course intelligent management system based on a large model, and belongs to the technical field of large models. Comprising a multi-modal data acquisition module, a privacy calculation preprocessing module, a dynamic knowledge enhancement module, a time sequence data analysis module, an intelligent decision engine module, a multidisciplinary collaboration module, a patient interaction platform module, a dynamic intervention feedback module and a system security center module. The cross-mechanism data security sharing is realized, and the compliance of sensitive information processing is also ensured; a two-channel medical knowledge base is constructed, authoritative guidelines can be synchronized, newest clinical research data can be analyzed in real time, the knowledge base is kept in the newest state all the time, and the frontier scientific basis is provided for clinical decisions; dynamic modeling and trend prediction are carried out on long-term monitoring data of a patient by adopting a hybrid neural network model, and potential health risks and development trends can be identified more accurately.

The invention relates to a charging pile intelligent operation management method and system based on the Internet of Things. The method comprises the following steps: performing Internet of Things topology construction on charging pile nodes to obtain a node communication distance matrix and a network connectivity matrix; performing deep learning dimension reduction and node screening based on the network connectivity matrix to obtain a fault feature vector; the fault feature vector and the charging pile operation state data are input into a hybrid neural network for risk calculation, a fault risk assessment value is obtained, and the hybrid neural network comprises a space perception sub-network and a time sequence analysis sub-network; performing normalization and dimension reduction mapping on the power response data of the charging pile to obtain a time domain mapping matrix; and performing particle swarm optimization calculation based on the time domain mapping matrix and the fault risk assessment value to obtain a charging power distribution scheme, and generating a control instruction set according to the charging power distribution scheme. According to the invention, the dynamic optimization of the charging power distribution scheme is realized, and the charging efficiency and the load balance are effectively balanced.

The invention provides a multi-feature fusion rumor detection method, system and device based on knowledge distillation, and mainly solves the problems that an existing model is high in calculation overhead, insufficient in feature fusion and insufficient in emotion utilization. The method comprises the steps of firstly obtaining multi-dimensional data such as social media original texts and comments; extracting deep semantic representation by using a pre-training model, and analyzing comment emotion features in combination with a hybrid neural network; then, features such as semantics, emotions, emoticons and populations are input into a hierarchical gating interactive fusion network (GIFN), and weights are dynamically adjusted to achieve effective fusion of multi-granularity features; in order to reduce complexity, a knowledge distillation framework is designed: a deep GIFN is used as a teacher network to generate a soft label, and a lightweight student network (LSTM) is guided to perform training. According to the trained student model, the parameter quantity is remarkably reduced, meanwhile, good detection performance is kept, the student model can be conveniently deployed in an actual content auditing system or edge equipment, and social content rumors can be efficiently recognized and judged.

The invention provides a thermal power plant thermal instrument fault diagnosis method and system based on vibration analysis, and relates to the technical field of fault diagnosis, and the method comprises the steps: collecting and processing a vibration signal of a thermal instrument, extracting an optimized vibration feature vector, constructing a hybrid neural network model, and training the hybrid neural network model to obtain an optimized fault diagnosis model; performing fault diagnosis on the vibration feature vector, and generating a fault type, possibility and confidence; and performing fault risk assessment and case reasoning, and generating a fault reason analysis report and a maintenance suggestion.

The present invention provides a wind power off-grid hydrogen production power supply topology and control method without a step-down transformer, including: obtaining wind speed data, using a wavelet threshold algorithm for denoising, predicting power based on an LSTM-GRU hybrid neural network, and extracting power fluctuation characteristics through multi-scale entropy analysis; obtaining load data, using Kalman filtering and recursive least squares to establish an impedance model, and extracting dynamic characteristics in combination with Hilbert transform; using a differential evolution algorithm for multi-objective optimization to generate an optimal switching sequence and power allocation strategy; achieving system control through space vector modulation, balanced control, soft switch control and harmonic compensation; achieving state estimation and disturbance suppression based on a sliding mode observer and a nonlinear observer; and achieving fault diagnosis and protection using a deep belief network. The present invention improves system efficiency and reliability and improves dynamic performance.

The invention discloses a rockburst risk early warning method based on TBM multi-source data fusion and a hybrid algorithm, and the method comprises the steps: collecting TBM tunneling parameters, geological exploration data and micro-seismic monitoring data of different rockburst levels, achieving feature complementation through multi-source data fusion, improving the completeness and reliability of early warning, constructing a hybrid neural network model based on CNN, LSTM and an attention mechanism, and carrying out the early warning of the rockburst risk. Feature weight distribution is optimized by introducing an attention mechanism, so that the model can adaptively focus key risk signals, and the applicability of model early warning is improved; and in combination with a fuzzy comprehensive evaluation algorithm and a Bayesian probability model, outputting four rockburst grades of no rockburst, slight rockburst, medium rockburst and strong rockburst and occurrence probabilities thereof, and realizing real-time dynamic early warning and probabilistic early warning of the rockburst grades. Compared with an existing method, the method has the advantages that the accuracy, timeliness and engineering applicability of rockburst early warning are remarkably improved, and real-time and accurate early warning of potential rockburst and the grade of the potential rockburst can be achieved.

The invention discloses a power equipment fault prediction method based on multi-modal data and related equipment, and belongs to the technical field of intelligent monitoring and fault prediction of power equipment, and the method comprises the steps: collecting and preprocessing the multi-modal operation data of the power equipment; extracting features of the preprocessed multi-modal operation data and performing global feature fusion to obtain multi-modal features; inputting the multi-modal features into a pre-trained fault prediction model to obtain a fault category and an occurrence probability of the power equipment; the fault prediction model is obtained by inputting the multi-modal features into a hybrid neural network for training. Multi-modal operation data of power equipment is collected, key features are extracted by using a deep learning model, multi-modal data fusion is performed, and a fault prediction model is constructed based on a hybrid neural network. Real-time fault early warning and maintenance suggestions are provided, the operation reliability of power equipment is improved, the unplanned shutdown risk is reduced, and the method is suitable for health management of transformer substations, high-voltage power transmission equipment and wind generating sets.

The invention discloses a personalized care method, device and equipment based on an AI technology and a storage medium, and solves the technical problem that an artificial intelligence dialogue system lacks dynamic memory, multi-dimensional user portrait updating and personalized care strategies in the prior art. The method comprises the steps of collecting data and preprocessing the data; the method comprises the following steps: generating a dynamic user feature vector by constructing an LSTM-GRU hybrid neural network structure; dynamic evolution of a user portrait is realized through incremental learning and memory attenuation, and a user multi-dimensional portrait is constructed; constructing a memory storage hierarchical architecture; extracting and optimizing keywords based on a medium-term memory storage system, and optimizing and constructing an interest graph by utilizing a hierarchical clustering method according to a keyword weight enhancement formula; based on a memory storage hierarchical architecture, establishing a causal relationship and a sequential relationship between entity nodes and emotion nodes to construct a memory association knowledge graph; and establishing a trigger mechanism, performing emotion memory composite analysis based on the memory association knowledge graph, and generating a care strategy. The method can be widely applied to the artificial intelligence technology field.

The invention discloses a multi-domain collaborative flexible load schedulable potential evaluation and energy management method, and relates to the technical field of park energy management, and the method comprises the steps: obtaining space-time multi-source data of a smart park, constructing a graph neural network power consumer clustering model based on iterative self-organization analysis, and generating a power consumption behavior portrait of a power consumer; based on power utilization parameters of building air conditioners, electric vehicles, park ponds and energy storage batteries in the smart park, the schedulable potential of the multi-element flexible resources is evaluated; and based on the hybrid neural network and the Harris eagle optimization algorithm, constructing a load prediction model, and predicting various types of energy loads in a future time period. The invention aims to establish an accurate model and method, accurately evaluate the schedulable potential of different types of flexible loads in different scenes, realize efficient utilization, energy conservation and emission reduction and optimal configuration of park energy, and improve the overall energy management level and operation efficiency of the park.

The invention provides a plateau lake agricultural non-point source pollution treatment method which comprises the following steps: acquiring vegetation indexes and surface temperature field data by using a remote sensing satellite, and generating a pollution source space thermodynamic diagram in combination with water quality and soil parameters of ground sampling points; performing space-time alignment and data fusion on the thermodynamic diagram and real-time runoff and soil permeability data acquired by the hydrological sensor network, and constructing a structured pollution migration database; on the basis of the database, a hybrid neural network model embedded with physical constraints is utilized to predict pollutant concentration distribution within 72 hours in the future; inputting the predicted value into a multi-stage optimization controller, and generating a control parameter set comprising treatment intensity, engineering parameters and a fertilization ratio; generating a treatment strategy map covering the drainage basin through a GIS; and deploying an Internet of Things monitoring node to collect the treated water quality data to form a closed-loop control link. The treatment efficiency and effect can be improved, the treatment cost is reduced, and the negative influence on the ecological environment is reduced.

The invention relates to the technical field of meteorological observation, and provides a ground hail identification method and system based on a hydrogel classification result, and the method comprises the steps: carrying out the time-space correlation of multi-source hail data through a time-space matching algorithm, and obtaining a hail event data set of time-space matching; through a dynamic membership function optimization algorithm, self-adaptive phase state identification is carried out on the dual-polarization radar data, and multi-elevation hail phase state characteristic parameters containing rain-ice mixture categories are obtained; based on the multi-elevation hail phase state characteristic parameters, performing integrated preprocessing on the multi-source meteorological data to obtain standardized multi-dimensional meteorological characteristic data fused with phase state characteristics; performing unsupervised pre-training and supervised fine-tuning training on the DCNN-DBN hybrid neural network through the standardized multi-dimensional meteorological feature data to obtain a ground hail recognition model; and outputting a hail falling area identification result through the ground hail identification model. According to the invention, the distinguishing capability of easily-confused phase states is improved, and the false alarm rate and the missing report rate of hail identification are reduced.

The invention relates to the technical field of collaborative optimization, in particular to a virtual power plant power generation-utilization-price collaborative optimization system based on an AI large model, and the system comprises a load confidence matching module, a resource stability mapping module, a source-load capacity coupling module, an electricity price interval adjustment module and a comprehensive regulation and control linkage module. According to the method, the confidence interval prediction of the load demand is realized based on the hybrid neural network modeling of the load behavior data and the equipment temperature control characteristic sequence, and the scheduling matching confidence is measured according to the boundary overlapping condition of the prediction interval and the power generation response characteristic; a stability screening mechanism for adjusting resources is constructed in combination with the output fluctuation ratio and the equipment inertia characteristic, the controllability of load adjustment and the real-time performance of source side response are improved, the price adjustment rhythm is corrected through an electricity price response delay factor, dynamic closed-loop linkage between load adjustment and price guidance is achieved, and the load adjustment efficiency is improved. The execution priority is dynamically updated under the condition that multiple response conditions are matched, and the certainty of resource scheduling and the sensitivity of response are improved.

The invention provides a wind power mixed tower damage prediction method and system based on digital twinning. The method comprises the steps of obtaining multi-source data of a wind power mixed tower; wherein the multi-source data comprises structure response data, environment load data and historical operation and maintenance data; based on a building information model and finite element model fusion technology, constructing a digital twinborn model of the wind power mixing tower; inputting the multi-source data into a digital twinborn model to realize real-time mapping of the wind power mixed tower and the digital twinborn model, and obtaining real-time structure state data of the wind power mixed tower through the digital twinborn model; inputting the multi-source data and the real-time structure state data into a pre-trained LSTM-GRU hybrid neural network model, and predicting a damage evolution trend of the wind power hybrid tower in combination with an attention mechanism; and on the basis of the damage evolution trend, a maintenance scheme of the wind power mixed tower is generated by using a multi-target particle swarm algorithm, so that balanced optimization of maintenance cost, power generation loss and risk level is realized, and the operation and maintenance efficiency and safety of the wind power mixed tower are integrally improved.

A drifting buoy trajectory prediction method based on a hybrid neural network prediction model, includes: S1, obtaining marine environmental data and historical trajectory data of a drifting buoy; S2, performing preprocessing on the marine environmental data and the historical trajectory data to obtain input data configured to predict northward and an eastward velocities of the drifting buoy; S3, inputting the input data into the hybrid neural network prediction model to obtain predicted values of the northward and eastward velocities; S4, calculating latitude and longitude coordinates of a trajectory point of the drifting buoy based on the predicted values of the eastward and northward velocities; and S5, predicting, by repeating the step S1-S4, latitude and longitude coordinates of trajectory points of the drifting buoy at multiple time points to obtain a sequence of trajectory point coordinates to thereby achieve trajectory prediction of the drifting buoy over a target future period.

The invention relates to the technical field of remote sensing hyperspectral images, in particular to a remote sensing hyperspectral image super-resolution reconstruction method based on a hybrid neural network, and the method comprises the steps: firstly carrying out the preprocessing of a remote sensing hyperspectral image data set, and obtaining a training set; constructing a hybrid neural network with a spectrum hypergraph branch, a space hypergraph branch, a semantic hypergraph branch, a sensitive wave band extraction branch and a recovery and reconstruction branch; training the hybrid neural network by using a training set to obtain a hybrid network model; and finally, inputting a to-be-reconstructed low-resolution remote sensing hyperspectral image into the hybrid network model to obtain a corresponding reconstructed remote sensing hyperspectral image. According to the method, space, spectrum and semantic feature information are fully combined, and a main body structure and detail change characteristics in an image are captured; the sensitive wave band extraction branch is provided to enhance cross guidance and fusion among multi-domain features, and the effect of super-resolution reconstruction is fully improved.

The invention relates to the field of wind speed and wind direction prediction, and discloses a wind speed and wind direction prediction method based on an improved TCN-LSTM, and the method comprises the following steps: collecting real-time monitoring multi-dimensional meteorological data of an offshore wind station, and carrying out the preprocessing; performing feature fusion and feature enhancement processing on the preprocessed data, wherein the feature enhancement processing comprises spatial feature extraction, tensor reconstruction and time sequence modeling; inputting the data after feature fusion and feature enhancement processing into a wind speed and wind direction prediction model based on an improved TCN-LSTM hybrid neural network, and predicting the wind speed and wind direction at a future moment; the wind speed and direction prediction model comprises a time-space separation TCN network, a multi-head self-attention mechanism, an adaptive Dropblock mechanism, an LSTM network and a Bayesian optimization module; according to the method disclosed by the invention, the accuracy, robustness and calculation efficiency of wind power prediction in a complex marine environment can be improved.

The invention discloses a big data prediction model intelligent ammonia spraying control method and system suitable for kiln flue gas denitration. According to the ammonia spraying control method, historical and real-time data of multiple parameters such as the fuel gas flow of the glass kiln are collected firstly, the NOx amount is predicted through modeling of the LSTM-GRU hybrid neural network, the ammonia spraying amount is calculated through self-adaptive control, and the accuracy is ensured through two-stage correction. And the calculated ammonia spraying amount is transmitted to a production master control system to adjust an ammonia supply valve, and automatic alarm and manual control are switched in case of failure. The system comprises a data acquisition module, an intelligent prediction module, a self-adaptive control module and the like, and further comprises a dynamic catalytic reaction simulation test platform for optimization. According to the invention, intelligent automatic control of the big data model is realized, the ammonia spraying amount is accurately adjusted, the denitration efficiency and the ammonia utilization rate are improved, the operation cost is reduced, and the system reliability and stability are enhanced.

The invention provides an AI-based assembly state intelligent monitoring method and system, and relates to the technical field of artificial intelligence, and the method comprises the steps: collecting a multi-angle real-time assembly process image of a target product, constructing an assembly digital twin model, inputting the image into a hybrid neural network in combination with an assembly knowledge graph, analyzing the assembly node coordinate and torque data deviation, and obtaining an assembly state of the target product. And obtaining an optimal assembly path and torque control parameters, and finally generating an assembly quality evaluation matrix. Through fault feature extraction and mode recognition, potential faults are predicted, and a dynamic risk coefficient is calculated. And when the fault occurrence probability exceeds a preset threshold value, the position and torque compensation parameters are calculated, an execution instruction sequence is generated to control the assembly execution mechanism to perform adjustment, closed-loop optimization control in the assembly process is achieved, the assembly precision and efficiency are improved, and the fault rate is reduced.

The invention discloses a scene text recognition method based on optimized multi-modal vision and language processing, which comprises the following steps of: firstly, normalizing image data; and then inputting the preprocessed data into the optimized visual model. The visual model extracts multi-scale space and semantic features through a convolution-Transform hybrid neural network, and enhances the feature expression ability by using a multi-scale attention mechanism; the language model corrects character probability vectors output by the visual model, and learnable position codes are introduced to optimize representation of the features. Through designing a bidirectional multi-modal interaction module, visual and language features are fused, and a self-adaptive fusion mechanism is used to generate high-quality multi-modal joint feature representation. In the application stage, the optimized model is deployed through an efficient reasoning framework, and the speed and accuracy of scene text recognition are remarkably improved.

The invention discloses a sensing node communication system and method with a self-healing mechanism. The sensing node communication system comprises a node detection module, a node communication module, a preprocessing module, a mixed training module and a self-healing control module. The system realizes periodic acquisition of multi-dimensional environmental parameters by configuring a plurality of sensing nodes, shares state information based on a self-organizing network, and constructs a local network health state map. The preprocessed data is used for training a hybrid neural network, and long-term dependency features and local abnormal fluctuation features of node states are extracted from the data to assess fault risks. When the risk score exceeds a threshold value, the system automatically executes self-healing operations such as dynamic route reconstruction and standby link activation. According to the invention, the stability and the adaptive capability of the underwater sensor network are improved, and the method has wide application value.

The invention provides a cellular network flow space-time prediction method and system based on a hybrid neural network, and belongs to the technical field of intelligent communication. The method adopts a layered deep neural network architecture, and comprises a data embedding layer, a space-time coding layer, a feature fusion layer and an output layer. The data embedding layer maps a historical traffic sequence, cross-domain external data and metadata into high-dimensional features; the space-time coding layer is used for respectively fusing one-dimensional causal convolution and a Mama neural network to extract multi-scale time features and densely connecting convolution and a multi-head attention mechanism to capture multi-scale space features through time and space modeling branches; the feature fusion layer realizes adaptive weighted fusion of spatial-temporal features, cross-domain features and metadata features by using a gating fusion mechanism; and the output layer performs linear transformation on the fusion features to generate a final prediction result. According to the method, the spatial-temporal dynamic capture of the service traffic is accurate, the prediction curve is highly fit with the true value, and the accurate prediction of the multi-service traffic of the cellular network is realized.

The invention discloses a corrugated carton production management system and method, and relates to the technical field of corrugated carton production, the system comprises a production scheduling module, a material tracing module, an equipment cooperation module, a multi-mode detection module and a cost control module, and solves the problem that supply and demand relationships, production efficiency and quality management and control are not coordinated; according to the technical key points, an order ID is coded into an integer chromosome, sequential crossover and reverse mutation operation are applied, a fuzzy comprehensive evaluation model is constructed by combining Hash chain table and local hill climbing method optimization, and factors are quantified through a trapezoidal membership function, so that the order delivery punctuality rate and the equipment utilization rate are improved, and the mold replacement cost is reduced; the real-time production cost is collected and calculated through Kafka and Flink, an operation cost method model is constructed to maintain a standard cost database, abnormity is detected through a CEP engine, the cost hyper-branched risk is predicted through an LSTM-Transform hybrid neural network, a thermodynamic diagram is generated through a cache result to assist decision making, the cost accounting delay time is shortened, and the hyper-branched processing efficiency is improved.

The invention relates to the technical field of chemical safety and intelligent control, in particular to an explosion-proof intelligent temperature control method and system for a hydrogen peroxide storage tank based on multi-modal monitoring, and the method comprises the steps: collecting multi-dimensional data such as temperature, pressure and H2O2 concentration through a sensor array, carrying out the Kalman filtering and wavelet transform denoising, recognizing local abnormal points through an improved WMDS-DBSCAN algorithm, and obtaining the explosion-proof intelligent temperature control of the hydrogen peroxide storage tank based on the WMDS-DBSCAN algorithm. Predicting a hot spot diffusion trend in combination with an unsteady heat conduction equation and an RF-ADI algorithm; weighted Delaunay triangulation is utilized to construct a self-adaptive monitoring grid, risk area probability distribution is output through an SAM-LSTM-CNN hybrid neural network, and a dynamic amplitude limiting fuzzy PID controller adjusts the opening degree of a nozzle to achieve temperature control. According to the method, multi-modal data and an intelligent algorithm are fused, abnormal accurate detection, risk prediction and self-adaptive temperature control are realized, a closed-loop optimization mechanism is formed, and the operation safety and reliability of the hydrogen peroxide storage tank are effectively improved.

Methods, systems, and apparatus, including computer programs encoded on a computer storage medium, for performing a multi-modal machine learning task using a neural network. In one aspect, a method comprises, receiving a request to perform a machine learning task on an input tuple comprising a first network input in a first modality and a second network input in a second modality; processing the first network input to generate a first embedded sequence; processing the second network input to generate a second embedded sequence; processing the first embedded sequence and the second embedded sequence using an attention neural network to generate an updated first embedded sequence and an updated second embedded sequence; and processing the updated first embedded sequence and the updated second embedded sequence to generate a final representation for the first network input and a final representation for the second network input.