CN-121981091-A - Multi-dimensional dynamic industrial data intelligent form generation method and system

CN121981091ACN 121981091 ACN121981091 ACN 121981091ACN-121981091-A

Abstract

The invention relates to the technical field of data intelligent processing, and particularly discloses a method and a system for generating a multi-dimensional dynamic industrial data intelligent form. The method comprises the steps of firstly solving the problem of space-time asynchronous dislocation of multisource industrial data through a mixed alignment algorithm of Beidou grid codes and a multi-mode event clock, generating a clean data pool, then utilizing an industrial chain hypergraph knowledge graph and a controlled random walk model to identify hidden fragile nodes based on clean data and real-time events and quantify multi-level interruption risks, and finally dynamically generating a decision form capable of intelligently adjusting clauses along with negotiation feedback through an adaptive mechanism integrating differential privacy and game theory according to a risk graph. The invention realizes the complete flow from the original data to the intelligent decision form, and can effectively cope with the challenges of data alignment, depth risk prediction and dynamic game negotiation in complex industrial environments.

Inventors

YANG YALI
XIE MIN
MA LINA
Niu Qunliang

Assignees

太原锦华广告有限公司

Dates

Publication Date: 20260505
Application Date: 20260126

Claims (10)

1.A multi-dimensional dynamic industrial data intelligent form generation method is characterized by comprising the following steps: s1, accessing multi-source heterogeneous industrial data streams from port operations, meteorological sensors and a market quotation platform in real time; S2, processing the multi-source heterogeneous industrial data stream by using a mixed alignment algorithm based on Beidou grid codes and a multi-mode event clock, eliminating space-time asynchronous data dislocation caused by transmission delay, sensor precision difference and man-made reporting errors, and outputting a clean data pool with space-time alignment; S3, using the space-time aligned clean data pool and external event intensity data acquired in real time as input, running a hidden fragile node identification model based on an industrial chain hypergraph knowledge graph and a controlled random walk algorithm, identifying hidden fragile nodes in the industrial chain, quantifying the probability of multi-stage interruption risk caused by the hidden fragile nodes, and outputting a multi-stage interruption risk graph with probability; S4, creating and outputting an intelligent form embedded with hidden heuristic clauses and a dynamic adjustment strategy according to the multi-level interruption risk map with probability and combining historical negotiation interaction data through a self-adaptive generation mechanism integrating differential privacy and incomplete information dynamic game reasoning, wherein the intelligent form can dynamically adjust core clauses according to feedback of a partner so as to guide negotiation trend risk hedging to be better solved.
2. A multi-dimensional dynamic industrial data intelligent form generation system, using the form generation method of claim 1, comprising: the multi-source asynchronous data acquisition module is used for accessing and caching real-time and batch industrial data from the physical sensor, the service system database and the external API interface in parallel; the space-time alignment and cleaning module is used for executing a mixed alignment algorithm in the second method step on the collected original data so as to solve administrative space-time dislocation and output a clean data pool; The hidden risk quantification deduction module is used for receiving the clean data pool and executing the identification model in the third step of the method to construct an industrial chain hypergraph knowledge graph and calculate hidden fragile node risk probability; a game self-adaptive form generation module which is used for generating and configuring a game clause policy space in a form template based on the risk probability map and executing a self-adaptive generation mechanism in the fourth method step; the differential privacy protection output module is used for adding noise protection to the sensitive risk data and the strategy parameters in the generated form and pushing the sensitive risk data and the strategy parameters to an external system; And the man-machine interaction feedback module is used for providing a configuration interface and capturing revision operation of the form by the user and negotiation feedback logs of the partner.
3. The intelligent multi-dimensional dynamic industrial data form generating system according to claim 2, wherein said multi-source asynchronous data acquisition module further comprises: A flow batch integrated access unit configured with a plurality of adapters to simultaneously process flow position data from the ship automatic identification system AIS, batch transaction data from the enterprise resource planning ERP system, and structured event data from an information API having some kind of characteristics in news; The data cache queue adopts a partition theme design to prioritize and temporarily store different data according to data sources and emergency degree; and the metadata management unit is responsible for marking each access data record with a label containing a data source identifier, an original timestamp, an expected timeliness level and a Schema version for a subsequent module to call.
4. The intelligent multi-dimensional dynamic industrial data form generation system of claim 3, wherein said space-time alignment and cleansing module further comprises: The Beidou grid coding unit is used for mapping each piece of industrial data containing geographic position information, such as port berthing events or warehouse entry and exit records, to a Beidou position code with global uniqueness through coordinate conversion, and taking the code as a main key for data aggregation and alignment in a space dimension; a multi-modal logic clock synchronization unit for maintaining an independent event logic clock for each entity or event, the advancement of which is not driven by absolute physical time, but rather triggered and calibrated by the occurrence of a series of critical business events associated with the entity.
5. The intelligent multi-dimensional dynamic industrial data form generation system of claim 4, wherein said space-time alignment and cleansing module further comprises: the asynchronous data fusion and conflict resolution unit receives the outputs from the Beidou grid coding unit and the multi-mode logic clock synchronization unit, sorts and aligns the data with the same Beidou grid coding in each cluster according to the time sequence of the logic clock, so that a plurality of records of the same service fact reported from different sources and different times are fused, the field value conflict occurring in the fusion process is automatically resolved by adopting a weighted voting algorithm based on the credibility weight and the time sequence freshness, and finally a clean data pool with consistent time and space and no conflict is output for a downstream module.
6. The system for generating a multi-dimensional dynamic industrial data intelligent form according to claim 5, wherein said implicit risk quantization deduction module further comprises: The industrial chain hypergraph knowledge graph construction unit is used for constructing an industrial chain hypergraph model with a node type comprising enterprises, infrastructures, logistics paths and policy regulations on the basis of entities and relations in the clean data pool, wherein the hyperedge represents a multi-element dependency relation; And the controlled random walk analysis unit is used for carrying out random walk on the hypergraph model by taking the abnormal data node identified in the clean data pool as a starting point, wherein the walk transition probability is dynamically modulated by external event intensity data acquired in real time and tends to traverse a path with low network topology redundancy, and finally, the hidden fragile node is positioned and the risk probability value of causing multistage interruption is calculated by analyzing the convergence condition and the weight accumulation of the walk path.
7. The intelligent form generating system of multi-dimensional dynamic industrial data according to claim 6, wherein the controlled random walk analysis unit is further internally provided with an event intensity modulation subunit and a path vulnerability assessment subunit, the event intensity modulation subunit analyzes information data with certain characteristics in news in real time through a natural language processing technology, extracts negative events related to specific regions, industries or entities and quantifies the intensity of the negative events, so as to dynamically adjust the probability of random walk-related superside transition, the path vulnerability assessment subunit is used for calculating topological redundancy indexes and dependency depths of each walk path, multiplying current event intensity modulation coefficients of each side of the path with business dependency weights to obtain the overall vulnerability score of a risk conducting chain represented by the path, and the overall vulnerability score is used as a basis for calculating probability values in an output risk map.
8. The system for generating a multi-dimensional dynamic industrial data intelligent form according to claim 7, wherein said game adaptive form generation module further comprises: The differential privacy reference form generating unit receives the risk probability map from the upstream, firstly applies Laplacian noise conforming to differential privacy definition to information related to specific risk values and internal evaluation parameters, and then fills a preset form template frame based on the noisy safety data to generate a reference form without revealing sensitive information; The game strategy dynamic embedding unit is responsible for automatically inserting a plurality of groups of tentative clause options with slight differences in the specific non-core clause positions of the reference form, and presetting a set of follow-up core clause adjustment strategy triggered according to feedback of a partner to different options to form a dynamic game strategy tree under incomplete information.
9. The system of claim 8, wherein the game adaptive form generation module is further connected to a negotiation feedback learning unit, and the negotiation feedback learning unit monitors the modification and acceptance of the tentative terms by the partner acquired through the electronic subscription system or the mail exchange in real time, uses the modification and acceptance as an action signal in the game process, and drives a lightweight game tree search algorithm to quickly find the current optimal response in the preset policy tree, so as to automatically adjust specific parameters or expression modes of core contents such as price terms, delivery cycle, default responsibility and the like in a subsequent revised version of the form, and realize dynamic adaptive optimization of the form contents to strive for risk hedging advantages in negotiation.
10. The intelligent form generating system of multi-dimensional dynamic industrial data according to claim 9, wherein said human-computer interaction feedback module comprises: the visual configuration platform provides visual display of the risk map, flexible editing of the form template and an adjustment interface of algorithm key parameters for service analysts; And the interaction log capturing unit is used for recording all manual coverage operations of the user on the system generation form, details of negotiating rounds with the partner and execution results of final signing in detail, and the logs are stored in a structured mode and fed back to the hidden risk quantification deduction module and the game self-adaptive form generation module, and are used for continuously optimizing the relation weight in the knowledge graph and the benefit function in the game strategy tree to form a complete system mechanism from data to form and then feeding back the data.

Description

Multi-dimensional dynamic industrial data intelligent form generation method and system Technical Field The invention relates to the technical field of data intelligent processing, in particular to a method and a system for generating a multi-dimensional dynamic industrial data intelligent form. Background The existing industrial data form generation technology is generally applicable to the conventional environments of single data dimension and stable service scene based on a preset static template and a rule engine. The application number CN202311546574.4 discloses a form acquisition system and a form generation method, wherein the system is a low-code form acquisition system, manual coding is reduced, development efficiency is improved, human errors and BUG are avoided, a user is helped to develop and maintain the form acquisition system more efficiently, a low-code development mode reduces a technical threshold, a deep programming technology is not needed, non-technical professionals can participate in development, development cost is reduced, manpower and time resources are saved, however, when complex scenes with wide data sources such as large-scale commodity cross-border trade, multi-mode intermodal logistics and the like and high real-time requirements and changeable environmental dynamics exist, the prior art has obvious limitations, the system is difficult to effectively process multi-source heterogeneous data from different administrative regions and conforming to different reporting systems and time references, so that the underlying data has serious space-time misplacement problems, subsequent analysis is built on an unreliable basis, the traditional risk assessment model is focused on directly associated upstream nodes, the hidden and non-linear relations in an industrial chain network are saved, the deep and non-dependency relation with the real-time performance is not predicted, the problem of the fact that the business system cannot be fully optimized according to the real-time performance is difficult to form the real-time performance of the real-time performance, and the risk is difficult to fill the real-time performance of the complex risk forms, and the real-time performance is difficult to fill the real-time performance of the real-time performance is difficult to optimally and the real-time conditions. Therefore, a new intelligent form generation scheme capable of systematically solving the above-mentioned layer-by-layer progressive problem is needed. Disclosure of Invention The invention provides a method and a system for generating a multi-dimensional dynamic industrial data intelligent form, which solve the problems in the background technology. In order to achieve the purpose, the invention is realized by the following technical scheme that the intelligent form generation method for the multi-dimensional dynamic industrial data comprises the following steps: s1, accessing multi-source heterogeneous industrial data streams from port operations, meteorological sensors and a market quotation platform in real time; S2, processing the multi-source heterogeneous industrial data stream by using a mixed alignment algorithm based on Beidou grid codes and a multi-mode event clock, eliminating space-time asynchronous data dislocation caused by transmission delay, sensor precision difference and man-made reporting errors, and outputting a clean data pool with space-time alignment; S3, using the space-time aligned clean data pool and external event intensity data acquired in real time as input, running a hidden fragile node identification model based on an industrial chain hypergraph knowledge graph and a controlled random walk algorithm, identifying hidden fragile nodes in the industrial chain, quantifying the probability of multi-stage interruption risk caused by the hidden fragile nodes, and outputting a multi-stage interruption risk graph with probability; S4, creating and outputting an intelligent form embedded with hidden heuristic clauses and a dynamic adjustment strategy according to the multi-level interruption risk map with probability and combining historical negotiation interaction data through a self-adaptive generation mechanism integrating differential privacy and incomplete information dynamic game reasoning, wherein the intelligent form can dynamically adjust core clauses according to feedback of a partner so as to guide negotiation trend risk hedging to be better solved. A multi-dimensional dynamic industrial data intelligent form generation system, comprising: the multi-source asynchronous data acquisition module is used for accessing and caching real-time and batch industrial data from the physical sensor, the service system database and the external API interface in parallel; the space-time alignment and cleaning module is used for executing a mixed alignment algorithm in the second method step on the collected original data so as to solve administrative space-time dislo