Search

CN-122001556-A - Industrial multi-mode data uplink intelligent agent implementation method based on communication module

CN122001556ACN 122001556 ACN122001556 ACN 122001556ACN-122001556-A

Abstract

The invention provides an industrial multi-mode data uplink intelligent agent implementation method based on a communication module, which is used for realizing wireless acquisition and access of multi-source sensing data of industrial equipment through the communication module and completing legal access control of the equipment based on an equipment identity authentication mechanism, combining natural language instructions and sensor data, carrying out semantic understanding and structural analysis on the industrial multi-mode data by utilizing a large model to form a unified multi-mode sensing result, judging task types according to the sensing result and carrying out intelligent scheduling, selecting corresponding data preprocessing and computing modes for different tasks to complete task processing, and carrying out structured packaging, digital signature and encryption processing on the analysis result and related data when the uplink condition is met, and submitting the analysis result and the related data to a blockchain network to complete storage. The invention realizes intelligent analysis and trusted uplink of industrial data and meets the requirements of multitasking and trusted sharing of data in complex industrial scenes.

Inventors

  • ZHAO ZHICHENG
  • MENG ZHU
  • HU SHIJING
  • ZHENG SIYANG
  • DING XIWEN

Assignees

  • 北京邮电大学

Dates

Publication Date
20260508
Application Date
20260212

Claims (9)

  1. 1. The method for realizing the industrial multi-mode data uplink agent based on the communication module is characterized by comprising the following steps: Step 1, industrial equipment D i collects multi-mode data X i generated by multi-source sensor data through a communication module C i , attaches equipment identity information ID i , and transmits the multi-mode data X i to a server S i ; Step 2, carrying out digital identity authentication on the industrial equipment according to the equipment identity information ID i , allowing the industrial equipment D i to be accessed into an edge intelligent agent if the industrial equipment passes through the digital identity authentication, and terminating a data interaction flow if the industrial equipment D i does not pass through the digital identity authentication; Step 3, acquiring sensor data X i and natural language input l i acquired by industrial equipment D i , and completing natural text semantic understanding and sensor data type analysis based on a large model to form a structured multi-mode sensing result F i ; Step 4, judging a task intelligent scheduling type T i according to the multi-mode sensing result, and automatically matching the processing flow based on the task type; Step 5, based on the current task type T i , generating a corresponding preprocessing operation sequence, and preprocessing the multi-mode data F i according to the preprocessing operation sequence; Step 6, performing task calculation processing on the preprocessed multi-mode data according to task processing requirements, and selecting basic feature processing or calling an artificial intelligent model to complete corresponding task processing according to task complexity; Step 7, sorting and expressing based on the task processing result to generate a structured analysis report containing a task conclusion, key characteristic information and task related contexts; and 8, if the current task relates to the industrial data uplink evidence-storing flow, sequentially carrying out structured packaging, digital signature and safe encryption on the analysis report and the preprocessed data, and generating a traceable evidence-storing certificate after block chain verification consensus.
  2. 2. The method as set forth in claim 1, wherein in step 1, the industrial device D i collects the multi-mode data X i generated by the multi-source sensor data through the communication module C i , and attaches the device identity information ID i to transmit to the server S i , specifically, the industrial device set may be represented as D i ={d 1 ,d 2 ,…,d n , any industrial device is configured with the communication module set C i ={c 1 ,c 2 ,…,c n corresponding thereto, and C i corresponds to D i one by one, and the communication module C i is connected to the server through a general purpose input/output interface (GPIO), The serial communication interface or the bus interface is connected with the industrial equipment D i and is used for collecting multisource information of environmental parameters, running states and physical quantities generated in the running process of the industrial equipment, and the communication module C i is used for carrying out time alignment on collected data, Data normalization and feature mapping processing to form corresponding multi-mode data representation X i ={x 1 ,x 2 ,…,x n , generating multi-mode data X i , embedding device identity information ID i uniquely corresponding to industrial device D i into data by a communication module C i , and combining data generation time t i to form a data packaging unit containing identity and a timestamp The data encapsulation unit is sent to the corresponding server S i through a wireless local area network or a cellular communication network, so that the centralized access, stable transmission and unified management of multi-mode data of the multi-industrial equipment are realized.
  3. 3. The method for implementing industrial multi-mode data uplink agent based on communication module as set forth in claim 1, wherein in step 2, digital identity authentication is performed on the industrial equipment according to equipment identity information ID i , specifically, matching verification is performed on the equipment identity information ID i and preconfigured equipment identity authorization information, when the matching result is passed, the corresponding industrial equipment D i is allowed to access an edge agent, a data channel for data interaction is established, so that data generated by the industrial equipment D i can be sent to the edge agent for processing, and when the matching result is not passed, the access request of the industrial equipment D i is refused, and the data interaction flow between the industrial equipment D i and the edge agent is terminated.
  4. 4. The method for implementing industrial multi-mode data uplink agent based on communication module set as set forth in claim 1, wherein in step 3, the sensor data X i collected by the industrial device D i and the natural language input l i are obtained, and natural text semantic understanding and sensor data type parsing are completed based on a large model, specifically, the large model performs semantic understanding on the natural language input l i to extract intention information and key semantic elements, and performs data type recognition and structure parsing on the sensor data X i to determine source type, data meaning and representation form of the sensor data, and based on this, alignment and fusion are performed on the natural language semantic understanding result and the sensor data parsing result to generate a structured multi-mode sensing result F i .
  5. 5. The method for implementing industrial multi-mode data uplink agent based on communication module set as set forth in claim 1, wherein in step 4, the task intelligent scheduling type T i is determined according to multi-mode sensing result, and based on task type automatic matching processing flow, specifically, the task intelligent scheduling type T i is taken from a predefined task type set T i ={T 1 ,T 2 ,T 3 ,T 4 ,T 5 , each task type corresponds to domain knowledge reasoning, network intrusion detection, report generation, fault diagnosis and data uplink, by comprehensively analyzing semantic information, data characteristics and context state contained in F i , determining task type T i to which the current task belongs, according to the task type T i , automatically matching corresponding processing flow, and scheduling the task to the corresponding processing module for execution, so as to complete processing operation adapted to the current task type.
  6. 6. The method for implementing industrial multi-mode data uplink agent based on communication module set as set forth in claim 1, wherein in step 5, a corresponding preprocessing operation sequence is generated based on the current task type T i , and the multi-mode data F i is preprocessed according to the preprocessing operation sequence, specifically, data cleaning is performed on the original data in F i , missing values, repeated values and invalid data items are removed, a cleaned data set f_clean i is obtained, time sequence processing is performed on the data set f_clean i in time stamp order, a time sequence representation F_clean i ={f_clean i (t 1 ),f_clean i (t 2 ),...,f_clean i (t n )}, is constructed to maintain continuity and consistency of the data in time dimension, abnormal value processing is performed on the time sequence data, abnormal data points are identified and corrected in a threshold judgment or statistical analysis mode, on the basis, characteristics related to the current task type T i are extracted from the processed data, characteristic vector V i ={v 1 ,v 2 ,...,v k is formed, characteristic transformation processing is performed on the characteristic vector V i to enhance the distinguishing and expressing capacity of the characteristics, the transformed characteristic is performed on the basis of the characteristic vector V i to meet the following dimension standard of the converted characteristic data, and the obtained after-dimensional characteristic standard is used for the subsequent standard processing of the data to obtain the numerical value representation 35.
  7. 7. The method for realizing industrial multi-mode data uplink agent based on communication module as set forth in claim 1, wherein in step 6, task computing processing is performed on the preprocessed multi-mode data according to task processing requirements, basic feature processing is selected according to task complexity or an artificial intelligent model is called to complete corresponding task processing, specifically, a task complexity index Q i is constructed and used for representing the requirement degree of the current task on computing capacity, semantic understanding depth and reasoning precision, wherein Q i is comprehensively determined by feature dimension scale, data time span and task semantic hierarchy factors, task complexity index Q i is compared with a preset complexity threshold value Qth, when Q i is smaller than or equal to Qth, the current task is judged to be a basic feature processing task, based on F i , basic feature processing operation of rule matching, threshold value judgment or statistical computation is directly performed, a corresponding task processing result R i is generated, when Q i > Qth, the current task is judged to be a high-complexity task, F i is called as an input to call depth learning model to perform task computing processing, and task processing is obtained through task type T i corresponding to task processing result R i .
  8. 8. The method for realizing industrial multi-mode data uplink agent based on the communication module as set forth in claim 1, wherein in step 7, the task processing result is arranged and expressed to generate a structured analysis report containing a task conclusion, key feature information and a task related context, specifically, the task processing result R i is arranged and expressed, specifically, the method comprises the steps of arranging and screening information in the task processing result, extracting core judgment results related to the task conclusion, key feature information and task related context information, organizing the information according to a preset report structure on the basis, and generating a structured analysis report, wherein the structured analysis report is represented by a data structure of a field or key value pair, including a task conclusion field, a key feature information field and a context information field, and the structured analysis report supports both machine-resolvable data format output and user-oriented readable form display for unified expression and output of the current task processing result.
  9. 9. The method for realizing the industrial multi-mode data uplink agent based on the communication module as claimed in claim 1, wherein in step 8, if the current task relates to an industrial data uplink evidence-storing process, the analysis report and the preprocessed data are sequentially subjected to structured packaging, digital signature and secure encryption processing, and the signed evidence-storing data is generated after block chain checking and consensus, specifically, the structured packaging is used for uniformly organizing content data of an evidence-storing object and corresponding metadata thereof, wherein the metadata comprises data source identification, generation timestamp, task type identification and processing state information to form a standardized data packaging format, the digital signature is used for carrying out signature calculation on the structured packaged evidence-storing data, the corresponding signature information is generated by introducing a private key signature mechanism to ensure the integrity and source credibility of the evidence-storing data in the generating and transmitting process, the secure encryption processing is used for carrying out encryption protection on the signed evidence-storing data to prevent unauthorized access or tampering in the transmitting and storing process, and completing the structured packaging, the digital signature and the secure processing, and the corresponding evidence-storing data are subjected to the block chain checking and the validity of the signed evidence-storing data, and the unique evidence-storing data can be recorded by the block chain after the block chain checking and the validity of the evidence-storing data is verified and the corresponding to the evidence-storing data.

Description

Industrial multi-mode data uplink intelligent agent implementation method based on communication module Technical Field The invention relates to the technical field of industrial Internet and artificial intelligence, in particular to an implementation method of an industrial multi-mode data uplink agent based on a communication module. Background With the rapid development of industrial internet and intelligent manufacturing, a large number of industrial devices continuously generate multi-source heterogeneous data including environmental parameters, running states, control information and the like in the production running process. In the existing industrial system, the data acquisition and transmission modes of industrial equipment mainly adopt wired connection or local and single communication modes, and the requirements of equipment mobility and remote access in complex industrial scenes are difficult to meet. On the premise of ensuring the real-time property and accuracy of the data, the method realizes the efficient acquisition and remote transmission of the data of the industrial equipment, and further performs unified processing and analysis on the multi-mode data generated by the industrial equipment, thereby realizing the comprehensive perception of the industrial operation state, and becoming a key problem to be solved in the intelligent development process of the industrial system. In the prior art, data acquired by industrial equipment are processed by relying on a single model, so that the combined understanding capability of natural language instructions and sensor data is lacked, and the method is difficult to adapt to complex and changeable industrial application scenes. In addition, the existing industrial data transmission and management mode still has defects in the aspects of data credibility evidence and responsibility tracing, and the requirements on data safety and credibility in industrial scenes are difficult to meet. Therefore, a method of uplink agent that can realize wireless access of industrial equipment data based on a communication module and perform intelligent processing on industrial multi-mode data and combine task scheduling and blockchain certification mechanism is needed to improve the intelligent level and data reliability guarantee capability of an industrial system. Disclosure of Invention The invention provides a method for realizing industrial multi-mode data uplink intelligent agent based on a communication module, which is characterized by comprising the following steps: Step 1, industrial equipment D i collects multi-mode data X i generated by multi-source sensor data through a communication module C i, attaches equipment identity information ID i, and transmits the multi-mode data X i to a server S i; Step 2, carrying out digital identity authentication on the industrial equipment according to the equipment identity information ID i, allowing the industrial equipment D i to be accessed into an edge intelligent agent if the industrial equipment passes through the digital identity authentication, and terminating a data interaction flow if the industrial equipment D i does not pass through the digital identity authentication; Step 3, acquiring sensor data X i and natural language input l i acquired by industrial equipment D i, and completing natural text semantic understanding and sensor data type analysis based on a large model to form a structured multi-mode sensing result F i; Step 4, judging a task intelligent scheduling type T i according to the multi-mode sensing result, and automatically matching the processing flow based on the task type; Step 5, based on the current task type T i, generating a corresponding preprocessing operation sequence, and preprocessing the multi-mode data F i according to the preprocessing operation sequence; Step 6, performing task calculation processing on the preprocessed multi-mode data according to task processing requirements, and selecting basic feature processing or calling an artificial intelligent model to complete corresponding task processing according to task complexity; Step 7, sorting and expressing based on the task processing result to generate a structured analysis report containing a task conclusion, key characteristic information and task related contexts; and 8, if the current task relates to the industrial data uplink evidence-storing flow, sequentially carrying out structured packaging, digital signature and safe encryption on the analysis report and the preprocessed data, and generating a traceable evidence-storing certificate after block chain verification consensus. Specifically, in step 1, the industrial device D i collects multi-mode data X i generated by the multi-source sensor data through the communication module C i, and attaches the device identity information ID i, and transmits the multi-mode data to the server S i, specifically, the industrial device set may be denoted as D i={d1,d2,…,dn, any in