2980 results about "Data interface" patented technology

The invention discloses a data sharing system and method for realizing multi-agent interaction, and belongs to the technical field of agents. According to the method, production command instructions are labeled, a large language model is trained, and task types are identified; disassembling the production command instruction into a subtask sequence; constructing a knowledge relationship graph through the graph database; establishing a unified data interface, and constructing a data sharing center; dividing agent types according to the power transmission and distribution service scene, and determining a function boundary of each agent; a multi-agent simulation environment is used for testing the cooperation efficiency of the subtasks, and the execution sequence of the subtask sequence is optimized; sub-tasks are dynamically allocated according to the intelligent agent load and the skill matching degree; the intelligent agent accesses the knowledge relation graph through a unified data interface, and real-time pushing of data change is achieved through a message queue; the decision of the large language model is optimized through cue word engineering, specific operation is executed by utilizing a tool library component, and the intelligent agent executes operation according to a subtask sequence.

The invention discloses a deep foundation pit stability deformation control method and device based on digital twinning. The method comprises the steps that monitoring time sequence data of a deep foundation pit are collected and obtained; constructing a deep foundation pit deformation control digital twin platform; constructing a CNN-LSTM composite neural network based on an Adam optimization algorithm through a foundation pit deformation prediction module, and performing model training / verification on the CNN-LSTM composite neural network through the data training set and the data verification set to obtain an optimized CNN-LSTM composite neural network; based on the optimized CNN-LSTM composite neural network, performing time sequence change prediction on the axial force of the steel support according to the data test set to obtain a prediction result; the steel support servo module realizes servo control on the axial force of the steel support according to a prediction result on the basis of a constructed servo control strategy; and the foundation pit BIM module is used for constructing a deep foundation pit BIM model based on a Revit modeling technology and interacting with the deep foundation pit BIM model in real time through a RevitAPI data interface according to the output of the steel support servo module so as to realize visualization of deformation control of the deep foundation pit. The problems that in the deep foundation pit construction process, due to the fact that complex geological conditions are involved, the surrounding environment and an existing building are affected to a certain degree, and deformation and stability of the deep foundation pit cannot be accurately controlled are solved.

The invention discloses a data acquisition method and system and a storage medium, and relates to the technical field of data acquisition and processing.According to the technical scheme, multiple sensor devices and data interfaces are integrated, multi-source data are acquired in real time, format standardization processing, time synchronization correction and spatial information alignment are carried out through a multi-mode fusion module, and the data acquisition efficiency is improved. Calculating to obtain a data consistency factor Tyhz and evaluating the data consistency factor Tyhz; when the data consistency factor Tyhz does not reach the standard, a data optimization module performs noise filtering and abnormal value elimination on a multi-source data set, and an intelligent analysis module calculates a risk prediction parameter Fcyz by using a convolutional neural network; the early warning evaluation module compares the Fcyz with a risk evaluation threshold Fth, calculates a risk early warning index Gyzs, compares the risk early warning index Gyzs with a risk early warning threshold E, and dynamically generates an early warning execution scheme, so that information pushing and emergency resource scheduling are realized, the problems of low multi-source data fusion efficiency and insufficient early warning precision are solved, and the safety of the system is improved. And the risk identification and emergency response capabilities of the urban emergency management system are effectively improved.

The invention provides an intelligent agent automatic generation and scheduling system based on an artificial intelligence large language model. The system comprises an intelligent agent configuration module driven by the large language model, a workflow engine and automation module and a front-end and rear-end integration module. The agent configuration module driven by the large language model realizes automatic configuration of an agent and a workflow based on the large language model, and comprises a large language model interaction layer and an agent configuration module; the workflow engine and automation module is responsible for execution logic of workflow, supports multi-task parallel processing and comprises a workflow engine module and a task scheduler; the front-end and rear-end integration module realizes cooperative work of a JavaScript front end and a Python Django rear end by designing a unified data interface, the front end is responsible for interacting with a user and displaying an execution state of a workflow, and the rear end provides service logic processing and data persistence.

The invention relates to the technical field of program exception positioning, in particular to an AI-based data interface program exception fault pre-positioning algorithm, which comprises the following steps: S1, data stream pre-analysis; s2, capturing an abnormal signal; s3, performing abnormal clustering and feature extraction; s4, interface behavior modeling is carried out; s5, abnormal traceability positioning; s6, generating a self-adaptive repair suggestion; s7, performing feedback circulation and continuous learning; and S8, performing visualization and report generation. Through the steps of data stream classification, key field extraction, behavior modeling and the like, the system can accurately capture abnormal signals in mass data, meanwhile, an abnormal mode is visually displayed through clustering analysis and a dimension reduction technology, a reliable foundation is laid for subsequent positioning and repairing, and in addition, the method has the advantages of being high in practicability and easy to popularize. The abnormal source is quickly positioned by adopting methods such as topological sorting and causal network reasoning, the false alarm rate and the missing report rate in the positioning process are effectively reduced by combining a sliding window and time sequence analysis, and the problem solving efficiency is greatly improved.

The invention relates to the technical field of intelligent bid evaluation, and provides an intelligent multi-dimensional bid evaluation analysis and decision-making method based on big data, which comprises the following steps: acquiring original bid evaluation data from a multi-source heterogeneous data interface, and fusing through semantic role labeling and a timestamp alignment algorithm to generate a time-space association data set. And performing multi-level cleaning to generate a high-confidence bid evaluation data set. And extracting a multi-dimensional index based on the domain knowledge graph, generating a dynamic feature tensor, and dynamically allocating a weight by adopting a coupling attenuation weight model. And constructing a bidder association network, calculating a node influence score, detecting a potential bidding behavior and generating a risk correction coefficient. And injecting the real-time data stream into the dynamic feature tensor, updating the index weight, and generating a three-dimensional scoring vector through a multi-target aggregation decision algorithm. And performing Pareto optimization by using the asymmetric game equilibrium model, and outputting an optimal bid-winning party sequence and a risk early warning report. The bid evaluation efficiency and fairness can be improved, and the bid invitation risk is reduced.

The invention relates to the technical field of intelligent control and unmanned aerial vehicle application, in particular to a bird repelling and unmanned aerial vehicle dispatching integrated platform based on big data analysis. Comprising a multi-source data acquisition module; a dynamic scheduling decision module; and a federal learning optimization module. According to the invention, the millimeter-wave radar, the infrared camera, the voiceprint recognition unit and the meteorological data interface are integrated through the multi-source data acquisition module, so that the defects of one-sided data and extensive evaluation of a traditional scheme are overcome; a bird threat level, unmanned aerial vehicle endurance time, an electric quantity attenuation model and an airspace control rule are fused through a dynamic scheduling decision module, dynamic balance of priority response of a high-risk area and unmanned aerial vehicle endurance and energy consumption is realized, and the problems of response lag and insufficient endurance caused by unreasonable task allocation in a traditional scheme are avoided; the federal learning optimization module adapts to a specific geographical environment through transfer learning, and the problem of strategy stiffness of a traditional scheme in a complex environment is solved.

The present invention relates to the field of building electromechanical engineering technology, and in particular to an intelligent comprehensive planning system for building electromechanical pipelines based on BIM and artificial intelligence. The technical solution adopted by the present invention is: including a multi-source data fusion module, building a data interface, seamlessly connecting CAD drawings, equipment parameter documents, construction site information and other multi-source data, and importing them into the BIM core database with one click; using a data cleaning algorithm to eliminate redundant and erroneous information, ensuring that the basic data of the model is accurate and reliable, and laying a solid foundation for subsequent pipeline planning; first, according to different types of buildings and different building spaces, classify and organize the electromechanical deepening design layout knowledge, and list all electromechanical systems and subsystems under the electromechanical system in the area. The advantages of the present invention are: it can reduce pipeline collisions by more than 90%, reduce rework costs by about 40%; shorten the construction period by more than 25%, and make construction resources more reasonably allocated; greatly reduce troubleshooting and maintenance time; and comprehensively improve the quality and benefits of building electromechanical engineering.

The invention relates to a building module data interface intelligent monitoring system based on the Internet of Things and a deployment method, and belongs to the technical field of building information monitoring. The system is composed of a distributed sensing unit, an edge computing gateway, a cloud platform and a visual terminal, a self-adaptive filtering algorithm and an abnormal mode recognition model are built in the edge computing gateway, the cloud platform adopts a time sequence database to construct a multi-dimensional data warehouse, and structural health degree evaluation is carried out by fusing an LSTM neural network and a random forest algorithm. The deployment method comprises the steps of optimizing a sensor distribution strategy based on a BIM model, establishing a wireless Mesh ad hoc network communication architecture, and configuring a grading early warning mechanism and a fault tracing function. The innovation point is that a dynamic threshold adjustment algorithm and an interface performance degradation prediction model are provided, and real-time monitoring of the connection state of the building module and life prediction are realized. The system has the advantages of flexible deployment, high detection precision and low maintenance cost, and the intelligent level of building structure safety monitoring is effectively improved.

The invention relates to the technical field of marine disaster monitoring and early warning, in particular to an AI-based marine disaster prevention monitoring and early warning system. Comprising a multi-source data acquisition module for acquiring multi-dimensional data through a sensor network, remote sensing data access and a meteorological data interface; the data preprocessing and fusion module is used for carrying out cleaning, space-time calibration and feature fusion on the collected data; the artificial intelligence analysis module is used for realizing accurate analysis and prediction of different levels of storm surge by using a multi-scale prediction model cluster unit; the risk assessment and early warning generation module is used for calculating risk indexes according to a related index system and dividing early warning grades; and the interaction and release module is used for realizing visual display and multi-channel release of the early warning information. According to the invention, advantages of multi-source data and artificial intelligence technology are integrated, and real-time and accurate monitoring and early warning can be carried out on storm surge and coast submerging conditions.

The invention relates to the technical field of intelligent control, in particular to an intelligent power optimization scheduling method in a new energy grid-connected scene, which comprises the following steps of: acquiring a current load, a rated load, an adjustment rate and a temperature rise value through a data interface, and normalizing to generate a characteristic parameter matrix; the product of the adjusting rate and the temperature rise value is used for calculating the correlation degree with the aging factor to obtain an elastic weight, extracting the voltage stability and the weight, inputting the voltage stability and the weight into a Dijkstra algorithm to screen a path meeting the safety margin, monitoring the slope of an inertia change curve, adjusting a compensation coefficient through fuzzy PID control, and feeding back the compensation coefficient to correlation degree calculation. According to the method, multi-source parameters are fused through a rule tree structure, the load regulation rate and the aging factor weight are dynamically distributed, bus voltage stability and elastic weight path optimization are combined, a synchronous inertia margin safety correction system is constructed, the system response precision, the state cooperation efficiency and robustness are improved, local optimum is avoided, and the system reliability is improved. And the standby capacity configuration and the frequency advance correction capability are optimized.

The invention provides an urban groundwater inrush risk assessment method based on a three-dimensional grid model and an XGBoost algorithm, and belongs to the technical field of geological disaster early warning. The method comprises the following steps: acquiring geological, environmental and historical disaster data in real time through a multi-source heterogeneous data interface; constructing a heterogeneous grid unit based on three-dimensional geological modeling, and dynamically binding parameters such as underground water level, fault activity and lithologic permeability coefficient; an XGBoost algorithm is combined with second-order Taylor expansion to optimize gradient calculation, and a nonlinear risk assessment model is constructed through Bayesian hyper-parameter tuning and SHAP value feature interpretive analysis; identifying contiguous dangerous areas based on spatial autocorrelation analysis and an adaptive threshold segmentation algorithm; and a risk thermodynamic diagram, geological section superposition and real-time early warning are realized through a 3D GIS platform. According to the method, the problems that a traditional method depends on a linear model, data is single and local risk description is insufficient are solved, and the accuracy, efficiency and engineering practicability of risk assessment are remarkably improved.

The invention discloses a unified storage management method for station data in expressway data, which comprises the following steps of: S1, constructing a multi-source heterogeneous data acquisition channel, and processing various data sources by using edge computing nodes to generate a standardized metadata packet; s2, establishing a dynamic fragment storage matrix, and performing three-dimensional slicing on the data from time, space and business dimensions; s3, designing an event-driven data association engine, establishing a space-time mapping relation, and realizing automatic association storage through event ID chain type binding; s4, deploying a multi-modal storage cluster, storing different data by using different storage units, and tracking data consanguinity through a distributed index service; s5, implementing dynamic vectorization authority management, calculating an access authority level according to cosine similarity, and establishing a fine-grained data access control model; and step S6, constructing a service data interface gateway. The method has remarkable advantages in the aspects of highway data acquisition, storage, association, management, sharing and the like.

The invention relates to an aviation equipment manufacturing digital main line engine system supporting multi-source heterogeneous data access, and belongs to the technical field of aviation manufacturing data management. The system comprises a multi-source heterogeneous data access adaptation module, a data fusion processing module and a full-process data link construction module. The multi-source heterogeneous data access adaptation module passes through a data interface protocol and format conversion component; the data standardization and fusion processing module completes association fusion through an aviation heterogeneous data collaborative computing framework by means of an aviation manufacturing field data element standard library, a unified heterogeneous data model is generated, and the framework comprises a data fragmentation layer, a parallel node layer and a result aggregation layer; and the full-process data link construction module constructs a data link covering the full life cycle of the product based on a time sequence association algorithm and a product unique identifier mapping mechanism. The system can effectively solve the integration and management problems of multi-source heterogeneous data in aviation equipment manufacturing, and improves the data processing efficiency and the whole-process data tracing capability.

The invention relates to the technical field of textile, in particular to a digital textile equipment management method based on big data, and the method comprises the following steps: 1, an equipment connection and data initialization stage: checking the data interface type of equipment in each link of a textile process, and connecting the data interface type to a workshop data collection network, establishing an electronic file for each device, setting initial data acquisition parameters, and carrying out pre-operation test and calibration on the devices; 2, a production process monitoring and data recording stage; according to the scheme, the equipment performance evaluation model is established by using a big data analysis technology, multiple factors such as energy consumption efficiency, a production stability index and an equipment failure rate trend are comprehensively considered, and the equipment performance is comprehensively and accurately evaluated, so that sudden equipment failures are effectively avoided, the risk of production interruption is reduced, and the production efficiency is improved. The overall management efficiency and the operation stability of the equipment are improved, the continuity of the production process is guaranteed, and the cost increase caused by equipment faults is reduced.

The invention relates to a credit line dynamic management system and method based on an AI large model, and the system comprises the following modules: a data interface module which is used for collecting the real-time financial data, transaction records, credit records and other related information of an individual or an enterprise; the multi-dimensional data analysis engine inputs the data into an analysis engine based on an AI large model; according to the analysis result, the AI model can automatically generate a personalized credit line and repayment deadline strategy for each borrower; and the market risk prediction model integrates the market risk model and predicts the influence of economic fluctuation on the borrowers. According to the invention, banks and financial institutions can provide more personalized and flexible credit services, the credit safety is ensured, and the service quality and the customer experience are improved.

The invention discloses an advertisement putting strategy optimization method and system based on preference data collaboration, and relates to the technical field of advertisement putting, and the method comprises the steps: collecting the internal user data of a local advertisement putting platform, obtaining the external user data of a third-party platform through a safety data interface, carrying out the integration and preprocessing of the data, and obtaining an advertisement putting strategy; extracting explicit preference data and implicit preference data; a user-advertisement scoring matrix is constructed based on the preference data, and user preference vectors are generated after decomposition; acquiring copywriting text, image and category information of an advertisement to be put, constructing an advertisement feature expression model and generating an advertisement feature vector; calculating a matching degree score based on the user preference vector and the advertisement characteristic vector, and dynamically optimizing an advertisement putting strategy according to the score; according to the method, through comprehensive utilization of explicit and implicit preferences of the user, more accurate user interest modeling and advertisement matching are realized, and the advertisement putting effect and putting efficiency are effectively improved.

The invention relates to the field of group quality management, in particular to a group quality management system and method based on a dynamic data source. Comprising a data source management module, a quality data acquisition module, a quality data storage and management module, a quality data analysis and statistics module, a quality control and decision support module, a user management and authority control module and a system setting and management module which are connected through data interfaces, the data flow, the use frequency and the updating condition are monitored, and an early warning mechanism for data source connection abnormity, data access timeout and data conflict is arranged. According to the method, data integration and sharing can be improved, the accuracy of data quality is improved, user experience is improved, and safety is enhanced. According to the method, quality abnormity and fluctuation can be quickly and accurately found, targeted control measures are taken in time, quality problems are restrained in the bud state, and defective products and waste products are effectively reduced.

A method for carrying out RAG index evaluation on an RAG system put into production comprises the steps that a test set is prepared in advance, and the content of the test set is lt; an original question and a correct answer gt; a two-tuple list; the RAG index evaluation method specifically comprises the following steps of 1, calling an RAG module interface of a production environment by using an original problem in a test set, and obtaining a model reply and a reference text list; 2, the original question, the model reply, the reference text list and the correct answer form evaluation data according to an input parameter format of an evaluation framework; and 3, calling an evaluation data interface of an evaluation framework by using the evaluation data obtained in the step 2, and generating an evaluation result by the evaluation framework. According to the method, RAG evaluation can be carried out under the condition that a model and a database interface used in a production environment are not provided.

The invention relates to the technical field of data storage, and discloses a lossless compression algorithm-based data storage method, which comprises the following steps of: receiving to-be-stored data through a data interface, automatically identifying a data format by using a data analysis tool, and performing feature analysis to obtain the size, the type and the complexity of the data; through deep analysis of data features, the method can flexibly select a proper Huffman coding processing mode, and optimizes a compression process according to different data types and complexity. In addition, a context modeling optimization mechanism is introduced, the coding mode is dynamically adjusted, and the compression ratio and the storage efficiency are further improved; efficient storage and reliability verification of data are ensured through a block storage strategy in combination with metadata management; the problem that an existing compression algorithm is low in efficiency is solved, the integrity and reliability of data storage are improved, and the method is suitable for storage requirements of large-scale data and diversified data.

The present disclosure describes methods, systems, apparatus, and media for object identification and classification, utilizing multi-feature and multi-modal data. This includes shape, material, brand, price, odor, taste, tactility, and sound. The system integrates a server space for data processing, a querying device for iterative searches, and a data interface module for refining results. It features AI-driven image optimization, feature extraction, and pattern recognition, employing novel techniques for fusing multi-feature and multi-modal embeddings utilizing multi-head attention. Additionally, a linker module powered by two active learning with feedback loops AI models consolidates scattered data into a unified object information database. The system also employs novel AI algorithms for isolating the object of interest through a saliency map and semantic analysis, as well as for enhancing raw images with a GAN-autoencoder.

The invention provides a digital employee robot performance intelligent evaluation method and system, and relates to the technical field of data processing, and the method comprises the steps: carrying out the feature extraction of user interaction contents in all scenes; performing fluctuation detection, identifying a time period of abnormal fluctuation, and extracting user request content and robot response content in the time period; restoring the current session path based on a preset business process template, performing path consistency comparison in combination with a response rule of the robot under the path, and generating an abnormal node set if the path is interrupted or the response is offset; extracting a user request format, a robot response behavior and an external data calling state corresponding to each abnormal node, performing abnormal source analysis, calculating user input abnormity, and identifying an attribution probability of abnormity and data interface abnormity; according to the invention, autonomy and accuracy of robot performance intelligent evaluation are improved.

The invention discloses an online monitoring system for partial discharge of a main transformer. The online monitoring system specifically comprises a series of modules including sensors, detection equipment, a data acquisition module, a data processing module, a data transmission module and an online monitoring platform, wherein the sensors are reasonably arranged outside the transformer by adopting multi-band combination of ultrahigh frequency, high-frequency current, ultrasonic waves and the like; the detection equipment realizes high efficiency and flexibility through a C language underlying driver and an assembly language data interface, and has various anti-interference and fault early warning technologies; the data acquisition module is used for conditioning an original signal and improving the data quality by adopting various filters; reliable and real-time data transmission is ensured by adopting a TCP / IP protocol; the data processing module comprises a time domain and frequency domain conversion sub-module and other sub-modules for auxiliary diagnosis; the online monitoring platform is based on Web, and is rich in function and extensible; the system monitors a plurality of main transformers in a transformer substation, collects signals through sensors, processes the signals through the modules and transmits the signals to a platform, meanwhile, the system is subjected to regular inspection, software and hardware upgrading optimization and the like to guarantee stable operation of the system, and support is provided for safety of an electric power system.

The invention discloses a multi-agent learning event flow scheduling method and device and a storage medium. The method comprises the steps that structured data and unstructured texts of an event text set are received through a preset data interface; extracting text semantic features of the event text set by adopting a preset text understanding model, and generating event text features from the extracted text semantic features through a structured data processing model; and inputting the event text features into an agent combination for data processing, and scheduling process nodes according to an output result of the agent combination. According to the method, a multi-modal data fusion mechanism and a self-adaptive process scheduling system are constructed, so that comprehensive extraction of event text features and efficient collaboration among intelligent agents are realized, the automation level and the analysis depth of interaction abnormal behavior event processing are remarkably improved, and more accurate and more efficient technical support is provided for event processing.

The invention relates to the field of heat supply heat load operation adjustment, and provides a continuous optimization heat supply load prediction method and system based on big data analysis learning, and the method comprises the steps: S1, configuring a data interface on a real-time library service module to receive real-time and historical operation data, the operation data comprises meteorological information data, heat exchange station heat load and room temperature qualification rate; and S2, preprocessing the collected operation data through a data processing module integrated on the heat supply platform service module, and analyzing the preprocessed operation data through an analysis algorithm stored in a relation module to determine a predicted heat load value of a next control period. The heat supply platform is combined, the big data analysis technology is used for learning and analyzing historical heat supply network operation data, proper correction and adjustment are carried out, the heat load of the heat supply system is effectively predicted, the method is closely combined with actual operation, prediction accuracy is remarkably improved, the operation process is simplified, energy use is optimized, and the operation efficiency and the system response speed are improved.

The invention discloses an active real-time interaction system based on LLM. The active real-time interaction system comprises a self-cognition module, a real-time perception module, an active planning module, a situation decision module, a dynamic execution module, an emotion engine module, an adaptive interaction module and a safety management and control module. The system adopts a multi-dimensional vector to represent role attributes, and constructs a hierarchical memory storage structure to record interaction experience; generating an environment state vector through a multi-modal data processing technology; executing the target task decomposition algorithm to generate a multi-path execution plan; making a real-time decision based on the multi-dimensional decision factor; mapping the abstract decision into an instruction sequence and monitoring an execution process; simulating a system emotional state and influencing decision expression; dynamically adjusting an interaction strategy according to the user characteristics; evaluating decision rationality and starting a corresponding intervention mechanism. All the modules form a closed-loop workflow through a standardized data interface, the problems that a traditional AI system is passive in response, lacks self-cognition, is limited in environmental perception, is single in planning capability, lacks flexibility in decision making and the like are solved, and active service, self-evolution and safety controllability of the system are achieved.

The invention relates to the technical field of commodity trade management, and discloses a commodity trade business management system based on a block chain. A block chain data acquisition module of the system obtains commodity basic information, trade enterprise qualification information and logistics node information in real time through a preset data interface, and synchronizes the information to a block chain distributed account book; the trade risk quantitative analysis module generates a trade link real-time risk coefficient set based on account book real-time data in combination with a trade compliance rule base, historical order fulfillment data and current logistics state data; a risk control strategy collaborative analysis module generates a plurality of candidate risk control strategy schemes when the risk coefficient exceeds a preset threshold value, calculates a collaborative execution cost factor when the enterprise qualification state is changed in combination with logistics node information, and obtains an interference influence value set of each scheme on the trade process; and a risk control strategy execution feedback module updates a block chain smart contract execution instruction according to the optimal risk control strategy scheme, and facilitates efficient management and control of commodity trade business.

The invention relates to the technical field of electric power science and technology project dynamic supervision, and discloses an electric power science and technology project dynamic supervision system based on a computer network. The power science and technology project dynamic supervision system based on the computer network comprises a data acquisition module, a data transmission module, a data storage and management module, a data processing and analysis module, a monitoring and alarm module, a user interface module, a safety and authority management module, a system integration module, a log and audit module and a maintenance and upgrade module. According to the electric power science and technology project dynamic supervision system based on the computer network, operation data of electric power equipment are collected in real time through the sensor network and the data interface, and the real-time performance and accuracy of the data are ensured in combination with efficient data transmission and storage technologies. The system supports intelligent data analysis and fault prediction, monitors the equipment state in real time and gives an alarm, so that the fault response speed and the management efficiency are remarkably improved, and the equipment damage and the power failure loss are reduced.

The invention relates to a stacking monitoring and excess material calculation method for multi-dimensional data linkage processing, and the method comprises the following steps: environment evaluation and demand definition: carrying out the evaluation of a stacking region, and carrying out the customization design of a sensor layout and a data collection scheme; sensor network deployment and calibration: deploying a sensor, an RFID tag reader and a 3D scanner according to a design scheme; building a data interface and a fusion platform; constructing a three-dimensional stacking model; development and testing of an intelligent analysis algorithm: developing the intelligent analysis algorithm based on machine learning, training and testing the algorithm, and optimizing the performance of the algorithm; a dynamic monitoring and anomaly detection mechanism is established, specifically, a dynamic monitoring system is implemented, stacking state changes are continuously tracked, an anomaly detection algorithm is developed, potential safety hazards are recognized in real time, and an early warning mechanism is triggered; excess material calculation and optimization suggestion generation: fine excess material calculation is executed, and accurate material excess information is provided; and integrating the system and debugging on site.

Embodiments described herein provide A method of training a neural network based model for predicting time series data. The method may include receiving, via a data interface, multi-variate time-series data; generating a plurality of tokens based on flattening the multi-variate time-series data; generating a first intermediate representation via a first cross-attention layer of the neural network based model with a plurality of dispatcher tokens as the query, and the plurality of tokens as the key and value; generating a second intermediate representation via a second cross-attention layer of the neural network based model with the plurality of tokens as the query, and the first intermediate representation as the key and value; generating a predicted time-series value based on the second intermediate representation; computing a loss based on a comparison of the predicted time-series value and a ground-truth value; and training the neural network based model based on the loss.