141 results about "Timing diagram" patented technology

The invention discloses a water supply pipe network water hammer control method and system based on multifunctional module fusion, and relates to the technical field of data identification. A time sequence diagram neural network prediction and traceability module which realizes water hammer risk prediction and propagation path traceability based on a causal constraint graph neural network model; the reinforcement learning intervention decision module is used for generating an active intervention strategy through a reinforcement learning agent, forming closed-loop control, abstracting a water supply pipe network into a graph structure, and modeling in combination with a time sequence, so that a propagation path of pressure waves can be comprehensively reflected, a blind area of traditional local modeling is overcome, and the comprehensiveness and accuracy of water hammer event detection are improved; and a causal analysis result is input as an adjacent matrix, so that the interference of irrelevant edges on prediction in topology is effectively eliminated, and the learning efficiency and causal traceability of the model are improved.

The invention discloses an APT attack path reconstruction method based on time sequence diagram comparison clustering and a medium. A security event standardized data set is obtained; mapping each security event into a multi-modal node through a heterogeneous time sequence diagram set construction method, and generating a directed edge to construct and complete a heterogeneous time sequence diagram set; a stage embedding time sequence diagram set is obtained through the joint attack stage set; outputting a time sequence diagram similarity matrix by adopting a multi-scale diagram similarity algorithm of time alignment perception; generating an event semantic sparse matrix based on the threat intelligence knowledge graph; obtaining an image clustering result set; uncertain samples in the graph clustering result set are obtained and processed, and an APT attack path reconstruction result is obtained. The problem that an APT attack path reconstruction method mainly depends on rule matching, single-dimensional feature comparison and manual analysis of security logs or alarm streams is solved, the interpretability of APT traceability is greatly enhanced, and subjective errors of manual research and judgment are reduced.

The invention provides a time-space dynamic graph construction method and system for an engineering construction project, and relates to the field of railway tunnel engineering management.The method comprises the steps that construction data entities are extracted, and an entity set is generated; taking entities in the entity set as nodes to be added into the graph, and configuring attributes for each node; adding a relation edge for each node according to a logic structure between the construction data; adding a time label and a space label to each node so as to form a data graph; associating each entity in the data atlas with a time chain and a space chain to form a data atlas attached to the time-space chain; combining a node merging strategy and a graph neural network to optimize a graph structure; and dynamically modeling dimensional changes of entities and relationships in time by means of a time sequence diagram neural network. According to the method, the dynamic change graph of the relationship between the construction data entities is constructed, so that the spatial-temporal information and the entity relationship of each process are accurately expressed and efficiently associated, and the retrieval efficiency and the management level are favorably improved.

The invention discloses an opportunity network link prediction method based on a time sequence diagram neural network, and particularly relates to the technical field of the time sequence diagram neural network, and the method comprises the following steps: S1, data collection and network snapshot generation; s2, weighting feature coding clusters; s3, GAT and LSTM combined processing embedding is carried out; s4, the probability is calculated through a multi-layer perceptron; s5, constructing a loss function according to the link; s6, drawing sets and training parameters; and S7, multi-index evaluation of the optimal model. The method accurately captures the spatial-temporal characteristics of nodes, collaboratively operates embedding, coding and decoding layers, improves the accuracy and stability of link prediction, stabilizes the reliability of network communication and the success rate of data transmission, innovates a loss function and an optimization strategy, solves the problems of data sparseness and over-fitting, enhances the generalization ability, adapts the model to a multi-field network, and improves the data transmission efficiency. And links are accurately predicted in scenes such as intelligent traffic and disaster rescue, resource configuration and communication efficiency are optimized, and innovative development of multi-industry network application is powerfully promoted.

The invention discloses a routing optimization method and system based on an attention mechanism and a time sequence diagram convolutional network, and belongs to the technical field of communication networks, the method classifies traffic in the network, when the network traffic is mouse traffic, a trained A3T-GCN model is adopted to predict the network state of each link, and when the network traffic is mouse traffic, the A3T-GCN model is adopted to predict the network state of each link; optimizing the link by adopting an improved butterfly optimization routing algorithm, calculating the time delay and the fluctuation of the link according to the network state, obtaining the link with the minimum time delay and the minimum time delay fluctuation through iterative optimization, and routing the mouse flow along the link corresponding to the optimal solution; when the network flow is an image flow, the A3T-GCN model is utilized to predict the average residual bandwidth of each link in the shortest hop count link set in the future w periods, and the link object flow with the maximum average residual bandwidth is adopted for routing; according to the method, distribution and utilization of network resources are optimized, so that the overall efficiency and throughput of the network are effectively improved.

The invention provides a risk conduction prediction method and system based on combined deduction of a time sequence diagram and a large model, and relates to the technical field of artificial intelligence and knowledge engineering, and the method comprises the steps: extracting a prospective time sequence fact based on a hierarchical cue word and a structured constraint mechanism; extracting conflicts and generating a sequential relationship; the historical time sequence knowledge graph is updated based on the time sequence relation and a confidence coefficient weighted updating strategy; determining a central entity, and initiating a structured query to the updated historical time sequence knowledge graph by taking the central entity as a starting point to generate a context knowledge sub-graph; converting the context knowledge sub-graph into a natural language description with a logic relationship, injecting a risk hypothesis event, constructing a cue word as a new input of a large model, and outputting to obtain a structured JSON object containing a complete reasoning chain; and converting the structured JSON object based on a risk path extraction and visualization algorithm influencing weight attenuation to obtain risk early warning information and a visual conduction path diagram.

The invention provides a three-dimensional modeling and construction management method and device and a readable storage medium. The method comprises the steps that multi-professional engineering drawings and unstructured texts related to construction design are acquired; extracting a geometric entity in each professional drawing in the multi-professional engineering drawings, and constructing a target three-dimensional skeleton model; a hierarchical attention mechanism and a time sequence diagram neural network T-GNN are adopted to extract semantic parameters from the unstructured text; fusing geometric data corresponding to the target three-dimensional skeleton model with the semantic parameters to generate a lightweight three-dimensional model; and for the complex conflict in the lightweight three-dimensional model, disassembling the complex conflict into a symbolic logic subtask and a neural network subtask. According to the method, the device and the readable storage medium, the problems of low modeling efficiency, insufficient semantic understanding and limited conflict optimization capability caused by excessive dependence of three-dimensional modeling of multi-specialty engineering drawings on a traditional BIM tool in traditional engineering design can be solved.

The embodiment of the invention discloses a static time sequence analysis method and device based on layout features and electronic equipment. According to the scheme, the gate-level netlist of the circuit can be obtained, the time sequence diagram is established according to the gate-level netlist, the nodes and the directed edges in the time sequence diagram are filled with the first statistical characteristics respectively, the first layout characteristics of each connection network in the time sequence diagram are calculated, and the time sequence diagram is expanded according to the first layout characteristics; and constructing a static time sequence analysis prediction model based on the graph neural network and the convolutional neural network, and predicting the delay of a connecting line time sequence arc, the delay of a logic gate time sequence arc and the delay of a pin node in the expanded time sequence diagram, thereby determining a static time sequence analysis result. According to the method, the layout characteristics of the circuit are obtained by extracting and coding the nodes of the connection network, and the static time sequence analysis prediction model is modeled according to the layout characteristics, so that the prediction accuracy of the prediction model in the layout areas with different aspect ratios is improved.

The invention provides a station group runoff prediction method based on a depth time sequence graph network, which is characterized in that a new station group runoff prediction model based on a graph neural network and a convolutional neural network is constructed for runoff prediction of multiple stations in a region, and a time sequence information extraction module is used for extracting feature information of input data on a time scale; the spatial aggregation module is used for aggregating feature information among different sites and describing the influence of other sites on the site; and the time sequence modeling part is responsible for performing time sequence modeling on information obtained by aggregation of each station, and finally performing full network connection on hidden state information to obtain predicted runoff of each station. According to the method, geographic space relations and mutual influence factors existing between different stations can be efficiently extracted, meanwhile, a traditional sequential network is fused, space-time relevance and consistency are comprehensively considered, hidden space-time relations in data are fully mined, and the accuracy of overall runoff prediction of the station group is improved.

The invention provides a calculation and power collaboration system and method, and belongs to the field of calculation and power collaboration, and the system comprises a data collection module which collects power data, calculation power data and price data, and constructs a power supply time sequence diagram and a calculation power demand time sequence diagram; a prediction module which generates a power supply predicted value and a computing power demand predicted value based on the power supply sequence diagram and the computing power demand sequence diagram; the intelligent matching module is used for generating a matching decision scheme based on the power supply predicted value and the computing power demand predicted value; the task scheduling module is used for performing differential scheduling based on the matching decision scheme; and the feedback optimization module is used for evaluating the task scheduling result and feeding back the evaluation result to the prediction module and the intelligent matching module. According to the method, the power supply and the computing power demand are predicted, the active scheduling decision is generated based on the prediction result, and the task scheduling module executes space-time flexible scheduling of the task, so that the consumption utilization rate of the new energy power and the supply stability of the computing power service are remarkably improved.

The invention relates to an open source component multi-modal dependency risk tracing method and device, and the method comprises the steps: employing a mode of combining static analysis and dynamic analysis to analyze a component dependency relationship of software, and combining AST analysis and construction script analysis; function-level features are extracted, a binary fingerprint database is constructed, the similarity between different versions is analyzed through LSH, behavior patterns in binary codes are analyzed, and hidden dependencies or malicious code injection is detected; the version updating history of the dependent component is analyzed and monitored by using a time sequence, and the vulnerability security of the component is evaluated in combination with attack graph analysis; high-risk components on the path are calculated, potential supply chain attack points are identified, and risk points are subjected to cross validation; a time sequence diagram database is used for recording the component dependency relationship, and time backtracking query is supported. According to the method, a multi-mode dependency analysis method is adopted, potential dependency risks are rapidly identified, and potential supply chain attack risks are timely warned.

The invention discloses a computing task scheduling method and system based on time sequence diagram network resource state prediction, and relates to the technical field of computing resource allocation, and the method comprises the steps: collecting computing power resource state data based on computing power nodes, and modeling the computing power resource state data into time sequence heterogeneous diagram data; constructing a time sequence heterogeneous graph attention network THGAT model, and performing offline training on the THGAT model by taking node feature and edge feature prediction loss as a loss function and taking minimization of a prediction error as a target; inputting the time sequence heterogeneous graph data into a THGAT model, and predicting the resource state of a future time window; the calculation task is converted into a calculation demand vector, and a multi-objective optimization resource allocation THGATRM algorithm is adopted to perform calculation power resource allocation on the calculation task according to the resource state of the corresponding time window; and executing the computing task according to the computing power resource allocation scheme. Through the technical scheme of the invention, the accuracy of computing power resource prediction is improved, and efficient allocation and optimization of resources are realized.

The invention provides a training method and system of a network community dynamic graph generation model and a graph generation method, and is applied to the technical field of data mining. The method comprises the following steps: acquiring a historical dynamic time sequence diagram of a target network, and carrying out step-by-step feature extraction on the historical dynamic time sequence diagram by utilizing a feature extraction module according to a time sequence to sequentially obtain a diagram embedding representation, a hidden variable and a hidden state; inputting the graph embedded representation into a community detection module to obtain a community detection result; inputting the hidden state into a community prediction module to obtain a community prediction result of the prediction time step; inputting the hidden variable and the hidden state of the previous time step into a topological structure generation module to obtain a generated network diagram; determining training loss based on the community detection result, the community prediction result of the prediction time step and the generated network diagram; and updating model parameters based on the training loss. According to the method, the problems that application scene requirements sensitive to dynamic changes cannot be met and the quality of generated dynamic graph data is not high in the prior art are solved.

The invention provides a novel traffic information dynamic time sequence diagram topological structure, which relates to the technical field of dynamic time sequence diagrams, and comprises the following steps: configuring a namespace and time granularity system of a traffic information dynamic time sequence diagram, setting a global timestamp rule and a propulsion strategy, defining an information identification bit structure, and setting a visibility and consistency control strategy. And a unified time window scheduling and snapshot life cycle management mechanism is configured. A dynamic extensible time sequence naming system is formed by establishing a unified naming space, time granularity delta t and a global timestamp rule, windowed organization and snapshot chain management of traffic information are achieved, the structure can automatically advance snapshot in a time driving mode and a data driving mode, time sequence consistency and real-time performance are guaranteed, and the time sequence naming efficiency is improved. The problems that node time management is disordered and data cannot be traced in a traditional traffic map model are solved.

The invention discloses a method for constructing vTEE and certification in an AMD SEV environment, provides a refined communication and fusing sequence diagram for the vTEE for the first time, and defines a module interaction sequence under two scenes of normal task execution and anomaly detection. According to the method, a trusted cloud security management platform initiates a request, a protection component coordinates a calculation component to collect an SEV-SNP report and vTEE key data, signature combination and uploading verification are carried out, and finally, the platform compares with a reference value to determine whether the vTEE is trusted or not. According to the method, a double-layer certification mechanism of SEV hardware certification and vTEE equipment measurement is adopted, an SEV-SNP certification report proves that a hardware environment, a VM mirror image and a vTEE operation environment are not tampered, vTEE key data measurement supports complete trust chain verification of equipment firmware, instruction streams and strategy states, and end-to-end certification is achieved. Equipment-level measurement, strategy issuing, abnormal fusing and remote certification are realized through an external protection part, and the system has important commercial landing value and strong system expandability.

The invention discloses an industrial firewall device based on time sequence atlas detection, which relates to the technical field of industrial control network security and comprises the steps of protocol identification, feature extraction, time sequence atlas construction, baseline modeling, online detection and response protection. Extracting time sequence characteristics such as a request-response relationship, a time interval and an event sequence, and converting the time sequence characteristics into a time sequence graph on which nodes correspond to communication events and edges correspond to time sequences; and generating a baseline map based on normal data, comparing the deviation between the current map and the baseline map in real-time communication, and executing a protection action after judging abnormality. The device for realizing the method comprises a data acquisition module, a feature extraction module, an atlas modeling module, a detection analysis module, a protection response module, a man-machine interaction module and a wide-voltage power supply module which are electrically connected in sequence. According to the method and the device, hidden time sequence attacks can be accurately identified, multiple industrial protocols are supported, and the adaptive resource-limited industrial equipment is designed in a lightweight manner.

The invention discloses a dynamic electronic target sequence intention recognition method and system based on a time sequence diagram structure, and the method comprises the steps: obtaining multi-moment target data for a target object which dynamically changes in a real-time scene; performing sequence division on the multi-moment target data by adopting an FP-Growth algorithm to obtain an electronic target sequence set; modeling and coding each electronic target sequence in the electronic target sequence set by adopting a heterogeneous time sequence diagram model to obtain a heterogeneous time sequence diagram sequence; according to the heterogeneous time sequence diagram sequence, time sequence dynamic hidden relation prediction and time sequence dynamic feature prediction are carried out respectively, and a hidden relation and prediction features are obtained; the hidden relation comprises hidden nodes and hidden edges; and according to the hidden relationship and the prediction features, constructing graph data at the next moment, and performing intention recognition through a graph attention convolutional network. The problem that the time-space incidence relation of the target behavior changing along with time is difficult to capture in the prior art is solved, and the prediction and intention recognition precision is improved.

The invention relates to a cross-modal video clip retrieval method based on a dual sequence diagram, which comprises the following steps of: giving an unedited video with any length and a text description, inputting the video and the text description into a cross-modal video clip retrieval model for processing, and finally obtaining a fusion activation diagram of the video and the text description, the video clip which is most matched with the text description is obtained by fusing the activation graph. The invention innovatively provides a semantic enhanced dual time sequence graph framework for text video clip retrieval. According to the method, visual content and semantic information are decoupled, and an enhanced visual appearance graph and an enhanced semantic graph are respectively constructed, so that the time sequence dependency relationship among video clips is more accurately captured. By means of the decoupling mode, the model can better distinguish fragments with similar visual content but different semantics.

The invention discloses an intelligent lock recognition method based on a time sequence diagram neural network, and the method comprises the following steps: collecting multi-source event data of an intelligent lock, unifying the format, embedding codes, and generating an event feature vector set; constructing a time-varying multi-relation graph, and executing message passing and feature aggregation to obtain a neighborhood aggregation result; constructing a dual-frequency memory time sequence diagram neural network, and extracting long-term and short-term behavior characteristics in slow-frequency and fast-frequency memory units; inputting the slow frequency memory state and the fast frequency memory state into a gating fusion layer, and combining time and scene information to generate a full-time-domain embedded vector; calculating identity matching, abnormal risk and forgery credibility, and outputting a comprehensive identification result; and executing instant response and security processing according to an identification result, and carrying out network updating. According to the method, the sequence diagram neural network and the double-frequency memory mechanism are utilized, multi-element dynamic recognition of the intelligent lock is achieved, and the method has the advantages of being high in self-learning, high in recognition accuracy and excellent in safety robustness.

The invention relates to a natural resource asset management and territorial space planning system fusion method and system. The method comprises the following steps: encoding territorial space planning vector data and natural resource asset registration data to generate a space-time binding code; constructing a spatial topological graph and a time sequence diagram based on the codes, respectively extracting planning constraint feature vectors and asset dynamic feature vectors, and fusing the vectors to generate a planning asset coupling scoring matrix; recognizing a conflict unit set based on the planning asset coupling scoring matrix; based on the conflict unit set and a preset business rule base, performing classification processing on each conflict unit in the conflict unit set to obtain a conflict type; and based on the conflict type, generating a service execution instruction set through a preset reinforcement learning decision model. According to the method, through space-time unified coding and feature fusion processing, the accuracy of data association and the scientificity of decision making are improved, so that collaborative governance of territorial space planning and natural resource asset management is realized.

The invention discloses an intelligent self-adaptive time sequence diagram analysis and structuring method. The method comprises the following steps: preprocessing an input time sequence diagram; identifying and extracting basic structure elements in the time sequence diagram; extracting text information in the time sequence diagram by using an optical character recognition technology; the recognized signal lines are analyzed, and signal state changes and key event points are extracted through a neural network; detecting arrows representing time sequence constraints, and performing arrow detection and classification by using a target detection network; establishing an association relationship between the graphic elements and the text; the association relationship and the identified elements are verified and corrected by using grammar rules and global layout information of the time sequence diagram, and unreasonable results are filtered; converting the time sequence relation model into a standardized structured data format; and the overall recognition performance is continuously optimized based on interactive correction and feedback of confidence. According to the method, the workload of manually converting the time sequence diagrams by engineers is greatly reduced, the accuracy of time sequence constraint extraction is improved, and various time sequence diagram formats and styles are supported.

The invention discloses a time sequence diagram neural network prediction interpretation method based on graph generation, which is an interpretation method taking a graph node sequence as input, does not depend on a data form input in specific time sequence diagram neural network prediction, and can interpret time sequence diagram neural network prediction based on snapshots and events at the same time. According to the method, the recurrent neural network (RNN) is utilized to construct the generative model, dynamic changes of nodes and edges in the time sequence diagram neural network are captured, and the method can adapt to input sequences of different lengths. According to the method, the pre-training interpretation generator model is adopted, the input graph node sequence is obtained through breadth-first search, the model convergence speed can be increased through the pre-training interpretation generator model, and meanwhile the interpretation accuracy is improved. The method has the advantages that the universality and practicability of the interpretation method are enhanced, and the calculation time of the interpretation generation model is shortened.

The invention discloses a power distribution network fault positioning method based on a time sequence diagram convolutional network, and the method comprises the following steps: 1), obtaining the time sequence measurement data of a power distribution network node, and carrying out the preprocessing, and obtaining a plurality of local time sequence segments; 2) performing feature extraction on the local time sequence segments by using a multi-layer one-dimensional convolutional network to obtain time sequence features; 3) processing the time sequence features by using a time sequence diagram neural network to obtain node features; and 4) fusing and mapping the features of the nodes by using a full connection layer, and outputting a fault positioning result. According to the method, the local feature mode and the global fault propagation dynamic state can be captured at the same time, the fault positioning accuracy of the power distribution network is greatly improved, and a solid technical support is provided for safe and stable operation of an intelligent power grid system.

The invention relates to the technical field of energy management, in particular to a multi-source heterogeneous energy load prediction method and system for a manufacturing enterprise cluster, and the method comprises the steps: collecting original energy data from a multi-source heterogeneous system in the manufacturing enterprise cluster and external environment information related to the energy load change, preprocessing the original energy data and the external environment information to obtain standardized energy time sequence data and final external environment information; based on the standardized energy time sequence data, the final external environment information and a pre-established equipment information model, constructing a triple heterogeneous atlas, and extracting structural semantic features of each node in the triple heterogeneous atlas; based on the structure semantic features, constructing a coupling sequence diagram, modeling the coupling sequence diagram, and identifying key factors; and inputting the structure semantic features and the key factors into a multi-scale prediction model, and outputting hourly load prediction curves of various energy media in a future prediction period.

The invention designs and implements a performance test environment construction method oriented to cross-chain transaction of block chains. The system supports a user to customize test scripts and execution parameters according to test requirements, tasks are executed and workloads are initiated at the same time through multiple actuators, a load scene of a tested cross-chain application system is created, the whole process of cross-chain transaction from an initiator to a receiver is tracked at the same time, key data are collected, and performance test indexes are calculated. And a visual result report is generated. Comprising the following steps: analyzing system requirements, and describing and analyzing functional requirements and non-functional requirements of a system; carrying out summary design on the system, and respectively designing and describing the system from a logic view angle, a process view angle, a development view angle and a physical view angle; detailed design is carried out on each module of the system, and a scheme is designed for the modules from two different perspectives of a class diagram and a time sequence diagram; and carrying out system testing and summarization on the system.

The invention relates to the technical field of computer system monitoring, and discloses a host process monitoring method which comprises the following steps: S1, multi-modal acquisition: acquiring process resources, behavior data and interaction relationships; s2, performing feature processing: performing normalization, time sequence modeling and interactive extraction; s3, behavior modeling: modeling a time sequence and an interaction graph to describe behaviors; s4, anomaly detection: identifying anomalies by time sequence and graph embedding and confrontation methods; s5, safety response: executing blocking, resource limiting and isolation measures; and S6, adaptive optimization: dynamically optimizing the model and the strategy by feeding back. Through file system analysis and network flow capture, process static information, dynamic behaviors and interaction relations are comprehensively collected, full life cycle data coverage is achieved, resource occupation is optimized in combination with a dynamic sampling strategy, efficient operation is achieved while data comprehensiveness and timeliness are improved, and the system performance is improved. And the balance between efficient operation and low resource consumption is achieved.

The invention discloses a QVT-O-based SysML static model conversion system, relates to the technical field of software development, and solves the technical problem that efficient and automatic conversion from a SysML static model to a dynamic model cannot be realized in the prior art. The system comprises an information extraction module which is used for analyzing and extracting necessary information from an obtained SysML static model, wherein the necessary information comprises the action and relation of key attributes in system dynamic behaviors; the rule definition module is used for defining a conversion rule based on QVT-O according to the necessary information, and determining a corresponding relation between elements in a static model of SysML and elements of a time sequence diagram in a dynamic model; the execution module is used for automatically converting the SysML static model into a dynamic model according to a conversion rule, and the dynamic model is expressed as a time sequence diagram; the effect verification module is used for verifying the consistency of the static model and the dynamic model; through automatic conversion and accurate dynamic model generation, the system development period can be remarkably shortened, repeated work is reduced, and the subsequent system maintenance and updating process is simplified.

The invention provides an unsupervised APT detection method based on a time sequence diagram attention network and a deviation network. The unsupervised APT detection method comprises the following steps: acquiring a system audit log of a fixed-width time window by using a sliding time window; entities related to the events and interaction relations between the entities are extracted, and a dynamic traceability graph is constructed in combination with occurrence time of the events; modeling the interaction behavior of each node adjacent to the node in the time dimension by using a time sequence diagram attention network so as to extract a node time dependent feature representation; predicting the type of an edge between adjacent node pairs by using a multi-layer perceptron to obtain a prediction error; evaluating the degree of deviation between the type of the predicted edge and the corresponding reference standard by using a deviation network; and based on the prediction error and the deviation degree, identifying the time window subjected to the APT attack and the corresponding attack chain. The method is remarkably superior to a traditional detection method in the aspects of detection precision and robustness, and an efficient and extensible solution is provided for APT detection.

The embodiment of the invention provides a weight confirmation method and device, equipment and a medium, and relates to the technical field of machine learning. According to the weight confirmation method, the obtained fund information of the target fund is input into the time sequence diagram neural network model, and the time sequence feature vector in the fund information is extracted through the time sequence convolutional network of the time sequence diagram neural network model, so that the sensitivity and prediction accuracy of fund position change can be improved; the adjacent feature vectors of the target fund and other fund are extracted through the graph attention network of the time sequence graph neural network model, the mutual influence between different fund can be captured, and the daily position weight is predicted through the prediction network of the time sequence graph neural network model in combination with the time sequence feature vectors and the adjacent feature vectors. The accuracy of daily position weight prediction can be improved. Furthermore, by using the predicted daily position weight, the investor can be prevented from missing the optimal transaction opportunity due to decision lag, so that the investment risk is reduced, and the rate of return on investment is improved.

The invention relates to a coastal fishery intelligent text and time sequence diagram generation algorithm and system, and aims to improve the accuracy and practicability of coastal fishery wind forecasting. According to the system, sea level average wind and horizontal air pressure gradient data of data sources such as EC-thin are acquired, and a fine wind power series grid forecasting result is generated by applying a grid forecasting calculation technology and a dynamic threshold setting method. Furthermore, the system judges the dominant wind level by using a Marking Square algorithm, and particularly performs deep learning KAN convolution algorithm correction for the wind level caused by typhoon so as to improve the forecasting precision. Finally, the system screens out wind power and wind direction data in a key period, automatically generates an intelligent text description and a time sequence diagram, and provides visual and accurate forecast information for fishermen and related mechanisms. The method has the beneficial effects that the forecasting accuracy is improved, the real-time performance and pertinence are enhanced, intelligent and visual output is realized, the emergency response capability is improved, and the sustainable development of fishery is promoted.