Rapid data acquisition method based on fusion of intelligent number asking and industrial connection technology

Abstract

The invention provides a rapid data acquisition method based on fusion of intelligent number asking and an industrial connection technology, and the method comprises the steps: obtaining industrial protocols of a plurality of devices and configuration parameters corresponding to the devices, and generating an acquisition model of each device; generating a structural fingerprint for each acquisition model; performing matching degree analysis on the plurality of acquisition models through multi-modal similarity calculation on the basis of the structure fingerprint, screening the plurality of acquisition models according to a matching degree analysis result to perform uniqueness management, and performing association mapping on the equipment and the corresponding acquisition model to obtain a de-duplicated acquisition model; edge gateway configuration accurate issuing of the collection model after duplicate removal is realized through a two-way transmission channel, and consistency is guaranteed by combining real-time structure fingerprint return and matching degree analysis. Through the AI-driven process arrangement and semantic understanding technology, the modeling efficiency, the information consistency and the issuing capability of the data acquisition system are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of industrial data acquisition, and particularly to a fast data acquisition method based on the integration of intelligent question answering and industrial connection technologies. Background Art

[0002] With the current digital transformation of the manufacturing industry facing severe challenges, due to the complex brand models of production equipment and inconsistent industrial communication protocol standards, there are significant differences in the configuration, entry, and sorting of industrial control protocols. Traditional data acquisition methods highly rely on manual operations. The entire process from protocol adaptation management to parameter distribution gateways needs to be manually completed by technicians, which is not only inefficient but also extremely prone to configuration errors. This backward manual management mode severely restricts the data acquisition efficiency, often causing delays in obtaining production data, and further affecting the timeliness of enterprise real-time monitoring, quality analysis, and decision-making optimization. There is an urgent need to achieve automatic protocol adaptation and rapid configuration distribution through intelligent technical means to improve the accuracy and timeliness of industrial data acquisition.

[0003] In the current field of industrial data acquisition, existing technical solutions mainly rely on manual configuration and traditional protocol parsing methods. This solution requires technicians to manually identify and judge the communication protocols of various industrial devices (such as PLCs, CNCs, sensors, etc.), and separately develop corresponding parsing and conversion modules for each protocol. In the data acquisition management part, it is necessary to manually create a data model of the device, define the mapping relationship of data points, and maintain a complex data point table. This process is not only time-consuming but also prone to configuration errors due to human negligence. In the edge gateway adaptation link, manual intervention is still required to configure communication parameters, protocol conversion rules, and data distribution strategies, making the entire data acquisition process inefficient and poorly scalable. In addition, due to the wide variety of manufacturing equipment and strong protocol heterogeneity, traditional solutions are difficult to achieve rapid deployment and flexible adjustment, severely restricting the large-scale application of the industrial Internet of Things.

[0004] The existing data acquisition solutions have the following main drawbacks and technical roots: 1. Low protocol adaptation efficiency. There are a large number of industrial protocols, and each protocol adopts different communication mechanisms such as register addressing methods and message structures. The existing technology lacks a unified protocol abstraction layer, and it is necessary to separately develop parsing codes for each protocol, resulting in a long development cycle and high maintenance costs.

[0005] 2. High manual configuration error rate. The device data model (such as point tables and variable mappings) depends on manual reading of device documents and then manual entry. Due to the lack of automated modeling tools, it is difficult to avoid human misreading of specifications or input errors, especially when facing complex devices (such as the DB block data of PLCs), the error rate increases significantly. Summary of the Invention

[0006] In view of the above situation, the main object of the present invention is to propose a fast data acquisition method and system based on the integration of intelligent question asking and industrial connection technology, so as to solve the problems of configuration complexity and low efficiency of manual modeling caused by differences in industrial control protocols in industrial data acquisition. Eliminate the specific development requirements for different industrial control protocols, and realize the automatic conversion and unified management of protocol configurations. Through automatic parsing of device documents, intelligent generation of device digital models and point tables is achieved, reducing manual intervention. Through an intelligent verification and conflict detection mechanism, the configuration and modeling error rate is controlled below 2%, significantly improving the reliability of data acquisition.

[0007] The present invention proposes a fast data acquisition method based on the integration of intelligent question asking and industrial connection technology, and the method includes the following steps: Step 1: Obtain the industrial protocols of several devices and the corresponding configuration parameters of the devices, and generate an acquisition model for each device through semantic parsing and structural conversion of the industrial protocols and configuration parameters; Step 2: Generate multi-dimensional fingerprint data based on the hash value, path structure and semantic information of the acquisition model, and use the multi-dimensional fingerprint data as the structural fingerprint of the corresponding acquisition model; Step 3: Based on the structural fingerprint, perform a matching degree analysis on several acquisition models through multi-modal similarity calculation to obtain a matching degree analysis result; Screen several acquisition models according to the matching degree analysis result for uniqueness management, and perform an association mapping between the devices and the corresponding acquisition models to obtain a deduplicated acquisition model; Step 4: Implement precise distribution of the edge gateway configuration for the deduplicated acquisition model through a bidirectional transmission channel, and combine real-time structural fingerprint feedback and matching degree analysis to ensure consistency.

[0008] The present invention significantly improves the modeling efficiency, information consistency and distribution ability of the data acquisition system through AI-driven process orchestration and semantic understanding technology around the three core capabilities of "intelligent identification of acquisition protocols, rapid generation of acquisition models, and automatic distribution of edge gateways".

[0009] Compared with the prior art, the beneficial effects of the present invention are as follows: 1. Application of intelligent question asking in the rapid implementation of edge data acquisition. The present invention is driven by the AI process orchestration engine based on FastGPT to guide intelligent question asking, realizes the import system of the original device acquisition point table with high differences in various industrial control protocols, automatically parses the field and structure relationships, generates a standard acquisition model, and the modeling configuration time of a single device is shortened from 1 hour of traditional manual operation to 5 minutes, and the overall data acquisition implementation efficiency is improved by more than 90%.

[0010] 2. Multiple algorithms of structural fingerprint and semantic embedding empower product model reuse. The present invention compares the similarity between device configuration models through dual algorithms of structural fingerprint and semantic embedding, realizes automatic identification and reuse suggestions of similar device models, avoids duplicate modeling, and can effectively improve the reuse rate of similar device models to over 98%.

[0011] 3. High coordination between the acquisition system and the edge gateway. The acquisition configuration information of heterogeneous devices can be sent to the edge gateway with one key and supports full / incremental synchronization. Combined with the gateway feedback ability, the system automatically compares fingerprints to determine whether the models are consistent or need to be redeployed, ensuring that the edge and center configurations are completely synchronized with an accuracy rate of 100%.

[0012] Additional aspects and advantages of the present invention will be given in part in the following description, become apparent in part from the following description, or be understood through the embodiments of the present invention. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flowchart of a fast data acquisition method based on the integration of intelligent data interrogation and industrial connection technology proposed by the present invention; Figure 2 is a logical tree structure diagram of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0014] Embodiments of the present invention will be described in detail below. Examples of the embodiments are shown in the drawings, where the same or similar reference numerals denote the same or similar elements or elements having the same or similar functions throughout. The embodiments described below with reference to the drawings are exemplary only for explaining the present invention and should not be construed as limiting the present invention.

[0015] These and other aspects of the embodiments of the present invention will become clear with reference to the following description and drawings. In these descriptions and drawings, some specific embodiments of the embodiments of the present invention are specifically disclosed to represent some ways of implementing the principles of the embodiments of the present invention, but it should be understood that the scope of the embodiments of the present invention is not limited thereto.

[0016] Please refer to Figure 1 and Figure 2 , this embodiment provides a fast data acquisition method based on the integration of intelligent data interrogation and industrial connection technology, and the method includes the following steps: Step 1. Obtain the industrial protocols of several devices and the corresponding configuration parameters of the devices, and generate an acquisition model for each device through semantic parsing and structural conversion of the industrial protocols and configuration parameters; As a preferred embodiment of the present invention, obtaining the industrial protocols of several devices and the corresponding configuration parameters of the devices, and generating an acquisition model for each device through semantic parsing and structural conversion of the industrial protocols and configuration parameters specifically includes the following steps: Obtain the tabular file data of configuration parameters, and verify the tabular file data to obtain the verified tabular file data; Use an AI model to perform semantic recognition on the column header names of the verified tabular file data, and combine with the industrial protocol dictionary library to determine the protocol parameter logic corresponding to each column, obtaining a structured data table with semantic tags; Clean the data of the structured data table with semantic tags, and infer the dependency rules between fields according to the column header semantics and the relationship between adjacent fields in the structured data table with semantic tags, obtaining a standardized configuration parameter table; Obtain the industrial protocol, and disassemble the standardized configuration parameter table into independent configuration units according to the type of the industrial protocol, obtaining a structured collection element set; Input the structured collection element set into a pre-trained task orchestration model (FastGPT) to generate configuration steps in logical order; Insert user confirmation nodes into the key steps in the configuration steps to obtain an interactive task flow diagram; Convert the interactive task flow diagram into a natural language Q&A process to guide the user to supplement missing parameters or correct logical conflicts, obtaining a complete set of configuration parameters confirmed by the user; Select a preset template according to the type of the industrial protocol, and fill the complete set of configuration parameters confirmed by the user according to the template structure to obtain the original configuration content; Uniformly convert the original configuration content into a format supported by the target system to generate a standard file with a version identifier, obtaining a collection model.

[0017] In the above solution, the present invention uses the AI process orchestration tool FastGPT to convert the "basic information for collection implementation" traditionally configured manually into an interactive intelligent question-and-answer process. By disassembling steps and generating task flows: Structurally decompose the collection elements (such as device addresses, register mappings, sampling periods, etc.) in industrial protocols such as Modbus and OPC-UA, and construct a task flow diagram through AI to generate a continuous interactive dialogue task.

[0018] And it also has the function of automatically recognizing and formatting the original Excel file; supports importing the original protocol information (such as an Excel table provided by the manufacturer), and the AI automatically parses based on the column header semantics and data structure, and recognizes the configuration meaning and parameter logic between fields; After the dialogue process is completed, the system automatically generates standard collection configuration content, outputs it in a unified Excel or JSON format, and can be directly docked with the data collection system to achieve one-key import.

[0019] Step 2: Generate multi-dimensional fingerprint data based on the hash value, path structure, and semantic information of the acquisition model, and use the multi-dimensional fingerprint data as the structure fingerprint of the corresponding acquisition model; As a preferred embodiment of the present invention, generating multi-dimensional fingerprint data based on the hash value, path structure, and semantic information of the acquisition model, and using the multi-dimensional fingerprint data as the structure fingerprint of the corresponding acquisition model specifically includes the following steps: Sort the basic information, attribute information, data type, and unit of each device in the acquisition model, and then calculate the hash value; Combine the fields in the acquisition model into a logical tree structure, and perform nested hierarchical encoding to obtain the structure path encoding; Use the AI semantic model to convert the device names and device attribute information in the acquisition model into semantic vectors; Combine the hash value, structure path encoding, and semantic vector into a unified structure to form multi-dimensional fingerprint data, and use the multi-dimensional fingerprint data as the structure fingerprint of the corresponding acquisition model.

[0020] Further, sorting the basic information, attribute information, data type, and unit of each device in the acquisition model and then calculating the hash value specifically includes the following steps: Perform field normalization processing on the basic information, attribute information, data type, and unit of each device in the acquisition model to obtain a standardized field string; Split the standardized field string into independent fields according to the delimiter, and sort them to obtain an ordered field sequence; Use a unified delimiter to connect the ordered field sequence in turn to obtain a spliced complete string; Apply an encryption hash algorithm to the spliced complete string to obtain the hash value of the device field.

[0021] Further, combining the fields in the acquisition model into a logical tree structure and performing nested hierarchical encoding to obtain the structure path encoding specifically includes the following steps: Combine the fields in the acquisition model into a logical tree structure, and use the fields of the nodes in the logical tree structure as the encoding result; Obtain the encoding results of all child nodes of the current node in the logical tree structure, and then sort them in lexicographical order to obtain the encoding set of the child nodes; Concatenate all the child node encodings in the encoding set of the child nodes in order into a single string, and use an encryption hash algorithm to obtain the hash value of the encoding; Take the first 4 digits of the hash value of the encoding as the unique identifier to obtain the aggregated hash value of the child node structure; Concatenate the fields of the node with the aggregated hash value of the sub-node structure. During the concatenation process, if the current node is not the root node, concatenate the parent node path separator to form a complete hierarchical path, and obtain the structure path encoding.

[0022] Further, using the AI semantic model, converting the device name and device attribute information in the acquisition model into semantic vectors specifically includes the following steps: Obtain the device name in the acquisition model and perform standardization processing on the device name to obtain the standardized device name; Obtain the device attribute information in the acquisition model, reorganize the device attribute information in the form of "attribute name: value + unit" to obtain a structured attribute description string; Concatenate the standardized device name and the structured attribute description string as the device complete description text sequence; Perform word segmentation on the device complete description text sequence and input it into the pre-trained BERT model, and use the hidden layer vector corresponding to the CLS flag bit output by the BERT model as the semantic representation of the overall text to obtain the semantic encoding vector of the text; Through a learnable industrial domain adaptation matrix, map the semantic encoding vector of the text to the industrial device parameter feature space and perform bias adjustment to obtain the domain-adapted semantic vector; Perform non-linear transformation and normalization operations on the domain-adapted semantic vector in sequence to obtain the final semantic vector.

[0023] In the above solution, in order to ensure the consistency and standardization of the acquisition configuration model, the system design uses the acquisition model structure feature fingerprint comparison algorithm to perform similarity matching and automatic association on the newly imported acquisition model and the existing models on the platform. The core process includes the following steps: Structural feature fingerprint generation. The system constructs structural features, that is, fingerprints, for each acquisition model. Each acquisition model finally forms a set of multi-dimensional fingerprint data consisting of field-level hash signatures + structural paths + semantic vectors, constituting its "structural identity". The specific content includes: 1. Field-level hash signature: Calculate the hash value after sorting the basic information, attribute information, data type, unit, etc. of each device; 2. Field structure path: Combine the fields in the model into a logical tree structure or path sequence, and perform nested hierarchical encoding on its structure path, such as Figure 2 shown; 3. Semantic embedding generation: Using the AI semantic model, convert the device name and device attribute information into semantic vectors, reflect the business semantic associations between different device models, and judge whether the device models are "of the same type and reusable". For example: Vector("1# Ball Mill") ≈ Vector("No. 2 Ball Mill"), Vector("1# Ball Mill. Rotation Speed") ≈ Vector("No. 2 Ball Mill. Main Shaft Rotation Speed"); Similarity comparison engine. The system determines whether a model is repeated or reusable by combining multi-dimensional feature similarity metrics. The main strategies are as follows: 1. Hash coincidence comparison: If the hash value of the newly constructed field has a coincidence degree higher than the set threshold (e.g., 85%) in the existing model, it is determined as a strong suspected repetition. 2. Structural path edit distance calculation: Calculate the edit distance (such as Levenshtein distance) for the structural path sequence to judge the similarity degree of the model organizational structure. 3. Semantic vector similarity comparison: Use cosine similarity to judge the semantic coincidence degree of the device model names to ensure the same meaning rather than just the same surface. 4. The comparison algorithm summarizes multiple indicators in a weighted fusion manner to form the final model similarity score and outputs the recommended operation (such as: reuse / overwrite / merge / new).

[0024] Step 3: Based on the structural fingerprint, perform a matching degree analysis on several collected models through multi-modal similarity calculation to obtain the matching degree analysis result; Screen several collected models according to the matching degree analysis result for uniqueness management, and perform an associated mapping between the device and the corresponding collected model to obtain the deduplicated collected model; As a preferred embodiment of the present invention, the specific steps of performing a matching degree analysis on several collected models through multi-modal similarity calculation based on the structural fingerprint are as follows: Establish a field mapping relationship for the multi-dimensional fingerprint data according to the field name or business logic for alignment to obtain a structurally aligned multi-modal feature data set; Traverse the hash value set in the structurally aligned multi-modal feature data set of all collected models, count the proportion of the same hash values, and use the proportion of the same hash values as the hash similarity score; Convert the structural path encoding in the structurally aligned multi-modal feature data set into a string sequence; Use the Levenshtein algorithm to calculate the minimum edit distance between the string sequences of all collected models and perform normalization to obtain the structural path similarity score; Project the semantic vectors in the structurally aligned multi-modal feature data set into a unified semantic space through a pre-trained semantic model; In the unified semantic space, calculate the cosine similarity of the same-name field vectors of all collected models to obtain the cosine similarity of all fields; Field weights of different magnitudes are assigned according to the field importance, and the cosine similarities of all fields are weighted and averaged using the field weights to obtain the semantic similarity score; Obtain the required business scenario types and the user's historical operation preferences, and dynamically generate multimodal weights according to the required business scenario types and the user's historical operation preferences to obtain the multimodal weight coefficients; Use the multimodal weight coefficients to perform weighted fusion on the hash similarity score, the structural path similarity score, and the semantic similarity score, and perform normalization to obtain the multimodal comprehensive similarity score.

[0025] As a preferred embodiment of the present invention, several acquisition models are screened according to the matching degree analysis results for uniqueness management. The specific steps for obtaining the deduplicated acquisition models are as follows: Divide the multimodal comprehensive similarity score into score intervals according to a preset rule and assign corresponding labels to obtain a list of candidate models with classification labels; among them, a value greater than or equal to 95% is used as the high matching interval and labeled as "strong reuse candidate"; a value between 70% and 95% is used as the gray area and labeled as "manual verification required"; a value less than 70% is used as the low matching interval and labeled as "independent new creation"; Obtain the set of acquisition models labeled as "strong reuse candidate" in the list of candidate models with classification labels, and perform uniqueness determination. If the same target model matches only one candidate model, directly associate and reuse it; If multiple candidate models match the same target acquisition model, they are dynamically sorted according to the acquisition model version time, the user's historical selection frequency, and the field integrity; the acquisition model with the highest priority is selected from the sorting result as the main version, and the rest are labeled as "historical copies"; and the fields missing in the main version are extracted from the "historical copies" to generate a patch file for manual confirmation to obtain the deduplicated set of main acquisition models; Obtain the set of acquisition models labeled as "manual verification required" in the list of candidate acquisition models with classification labels, and visualize the differences in the similar or conflicting parts to guide the user to correct and update the acquisition models labeled as "manual verification required" to obtain the corrected set of acquisition models; Associate and map the set of acquisition models labeled as "independent new creation", the deduplicated set of main acquisition models, and the corrected set of acquisition models with the corresponding devices to obtain the deduplicated acquisition models.

[0026] In the above solution, the present invention displays the standard acquisition configuration content of the dialogue process in the form of an online table, and marks the suggested operations according to the similarity for each row. The user makes the final confirmation. After confirmation, the model reuse association or the new model creation and association are automatically completed to establish a mapping relationship: If the similarity score is higher than the set threshold (95%), the system will determine that an existing model exists and recommend reuse; If the score is within the gray area range (70% - 95%), the system will highlight the different fields and recommend that the user confirm; If the score is within other ranges (less than 70%), the system will determine that there is no existing model and recommend that the user create and confirm; After the user finally confirms and completes, a mapping relationship is automatically established between the collection point and the original product model or the newly created product model. For the newly created product model, the model fingerprint will be automatically saved to support fast matching and mapping during the subsequent collection configuration process.

[0027] Step 4: Implement precise distribution of the edge gateway configuration for the deduplicated collection model through a bidirectional transmission channel, and combine real-time structure fingerprint feedback and matching degree analysis to ensure consistency.

[0028] As a preferred embodiment of the present invention, implementing precise distribution of the edge gateway configuration for the deduplicated collection model through a bidirectional transmission channel and combining real-time structure fingerprint feedback and matching degree analysis to ensure consistency specifically includes the following steps: Check whether the model fields cover all protocol parameters of the target device for integrity verification; based on the structure identity fingerprint, compare the models deployed on the target gateway for conflict pre-detection; After passing the integrity verification and conflict pre-detection, encapsulate the verified model into a configuration package with attached structure identity metadata according to the gateway protocol requirements; According to the configuration package size and field complexity, select different compression algorithms for compression processing, and select corresponding transmission rules according to the current network environment status to send the configuration package to the edge gateway; During the process of sending the configuration package to the edge gateway, the deployment status is real-time feedback through the bidirectional channel, and it is confirmed whether to enable incremental retransmission according to the deployment status to retransmit the conflicting fields to ensure the integrity of the collection model deployment and complete the collection model deployment; After the edge gateway completes the deployment of the collection model, obtain the actual effective configuration information of the edge gateway, and extract the field-level hash signature and structure path of the gateway configuration at a preset cycle to generate a new fingerprint structure; Compare the difference between the original fingerprint structure at the time of distribution and the new fingerprint structure, and mark the modified fields to obtain real-time structure identity data and difference marks; Obtain the different fields in the real-time structure identity data according to the difference marks, perform multi-modal similarity calculation on the different fields for incremental comparison; obtain a real-time consistency analysis report, and perform abnormal configuration repair and policy optimization based on the real-time consistency analysis report.

[0029] In the above solution, the present invention realizes efficient synchronization and consistency verification of data acquisition configuration information with the edge gateway, including: Unified collation and confirmation of configuration information The system displays the standard acquisition configuration content passed through the dialogue process in the form of an online table, and marks the recommended operations according to the similarity for each row, and the user makes the final confirmation.

[0030] Gateway issuing policy configuration and execution After confirming the configuration accuracy, the system supports synchronizing the acquisition model to the edge gateway as needed. The issuing mechanism includes the following key steps: A. Gateway information management. The system supports manually entering the connection information of the gateway through the gateway management function and testing the connection accuracy; B. Target gateway selection and configuration issuing. The user can check the target gateway in the interface, issue the JSON configuration in a unified structure format, and complete the edge push through the MQTT topic or API interface. The push process supports log echo and failure retry mechanisms. The system provides the following two issuing policies: Full issuance: Issue all the acquisition configurations of the current device at one time; Selected issuance: The user filters and issues some devices or acquisition points.

[0031] Gateway information feedback and consistency verification To prevent configuration drift between the edge gateway and the central system, the system supports the edge gateway to actively feedback the acquisition configuration information and perform the following operations: A. Generation of feedback field structure fingerprint Extract the field structure and device parameter information from the feedback content to generate a new structure feature fingerprint.

[0032] B. Comparison with the platform model fingerprint The system compares whether the feedback model fingerprint is consistent with the existing model in the system: Completely consistent: Confirm that the synchronization is successful and mark it as "deployed"; Fingerprints are close but inconsistent: Mark it as "the gateway-side configuration may have been manually modified" and prompt the user to confirm; Completely inconsistent: The system records the exception and prompts the user to redeploy.

[0033] C. Avoid duplicate mapping If the gateway manually configures the device information with the same structure in the existing model, the system can automatically determine and prevent duplicate mapping and duplicate issuance through fingerprint comparison, so as to keep the model clear and the data unique.

[0034] To verify the effectiveness of the present invention, taking the data acquisition of copper processing equipment as an example: The system automatically recognizes the S7 protocol of Siemens S7-1200 PLC and Modbus RTU pressure sensors in the production line, and uniformly maps the DB block address of the PLC and the sensor register address to standardized variables.

[0035] By parsing the device technical documents, the system automatically generates a digital model containing parameters such as temperature and pressure, and supplements metadata such as units and alarm thresholds based on the metallurgical industry template. The system automatically detects and corrects configuration conflicts. And when the system automatically detects a configuration conflict, such as a repeatedly defined register, it automatically corrects the configuration.

[0036] It should be understood that although the steps in the flowcharts of the embodiments of the present invention are shown in sequence according to the arrows, these steps are not necessarily executed in the order indicated by the arrows. Unless there is a clear indication in this article, there is no strict order restriction for the execution of these steps, and these steps can be executed in other orders. Moreover, at least a part of the steps in each embodiment may include multiple sub-steps or multiple stages. These sub-steps or stages are not necessarily executed at the same moment, but can be executed at different moments. The execution order of these sub-steps or stages is not necessarily sequential, but can be executed alternately or in turn with at least a part of other steps or sub-steps or stages of other steps.

[0037] It should be understood that each part of the present invention can be implemented by hardware, software, firmware or a combination thereof. In the above embodiments, multiple steps or methods can be implemented by software or firmware stored in a memory and executed by a suitable instruction execution system. For example, if implemented by hardware, as in another embodiment, any one or a combination of the following well-known technologies in the art can be used: discrete logic circuits with logic gate circuits for implementing logical functions on data signals, application-specific integrated circuits with appropriate combinational logic gate circuits, programmable gate arrays (PGAs), field programmable gate arrays (FPGAs), etc.

[0038] In the description of this specification, the description referring to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples.

[0039] The above-described embodiments merely represent several implementation manners of the present invention. The description is relatively specific and detailed, but it should not be construed as a limitation on the scope of the patent for the present invention. It should be noted that for those of ordinary skill in the art, without departing from the concept of the present invention, several modifications and improvements can still be made, and these all fall within the protection scope of the present invention. Therefore, the protection scope of the patent for the present invention shall be subject to the appended claims.

Claims

1. A fast data acquisition method based on the integration of intelligent question asking and industrial connection technology, characterized in that, The method comprises the following steps: Step 1: Obtain the industrial protocols of several devices and the configuration parameters corresponding to the devices, and generate a collection model for each device by semantically parsing and structurally converting the industrial protocols and configuration parameters; Step 2: Generate multi-dimensional fingerprint data according to the hash value, path structure and semantic information of the acquisition model, and use the multi-dimensional fingerprint data as the structural fingerprint of the corresponding acquisition model; Step 3: Based on the structural fingerprint, a matching degree analysis is performed on several acquisition models through multimodal similarity calculation to obtain a matching degree analysis result; According to the matching analysis results, several acquisition models are screened for uniqueness management, and the devices are associated and mapped with the corresponding acquisition models to obtain the deduplicated acquisition models; Step 4: The deduplicated collection model is accurately distributed to the edge gateway through a two-way transmission channel, combined with real-time structural fingerprint feedback and matching analysis to ensure consistency.

2. A fast data acquisition method based on the integration of intelligent question answering and industrial connection technology according to claim 1, characterized in that, In step 1, the industrial protocols of several devices and the configuration parameters corresponding to the devices are obtained, and the industrial protocols and configuration parameters are semantically parsed and structurally converted to generate a collection model for each device, which specifically includes the following steps: Acquire the table file data of the configuration parameters, and verify the table file data to obtain the verified table file data; The AI model is used to perform semantic recognition on the column header names of the inspected table file data, and the protocol parameter logic corresponding to each column is determined in combination with the industrial protocol dictionary library to obtain a structured data table with semantic labels. Perform data cleaning on structured data tables with semantic tags, and infer the dependency rules between fields based on the column header semantics and the relationship between adjacent fields in the structured data tables with semantic tags to obtain a standardized configuration parameter table; Obtain the industrial protocol, and decompose the standardized configuration parameter table into independent configuration units according to the type of the industrial protocol to obtain a set of structured collection elements; Input the structured collection element set into the pre-trained task orchestration model to generate configuration steps in a logical order; Insert user confirmation nodes into the key steps of the configuration step to obtain an interactive task flow diagram; Convert interactive task flow diagrams into natural language question-answering processes, guide users to supplement missing parameters or correct logical conflicts, and obtain a complete set of configuration parameters confirmed by users; Select a preset template according to the type of industrial protocol, fill the complete configuration parameter set confirmed by the user according to the template structure, and obtain the original configuration content; The original configuration content is uniformly converted into a format supported by the target system to generate a standard file with a version identifier to obtain the acquisition model.

3. A fast data acquisition method based on the integration of intelligent question number and industrial connection technology according to claim 2, characterized in that In step 2, generating multi-dimensional fingerprint data according to the hash value, path structure and semantic information of the acquisition model, and using the multi-dimensional fingerprint data as the structural fingerprint of the corresponding acquisition model specifically includes the following steps: The basic information, attribute information, data type, and unit of each device in the acquisition model are sorted and then the hash value is calculated; The fields in the collection model are combined into a logical tree structure and nested hierarchical encoding is performed to obtain a structural path encoding; Use AI semantic models to convert device names and device attribute information in the acquisition model into semantic vectors; Combine the hash value, structure path encoding, and semantic vector into a unified structure to form multi-dimensional fingerprint data, and use the multi-dimensional fingerprint data as the structure fingerprint of the corresponding acquisition model.

4. A rapid data acquisition method based on the integration of intelligent question number and industrial connection technology according to claim 3, characterized in that The steps for calculating the hash value after sorting the basic information, attribute information, data type, and unit of each device in the acquisition model are as follows: Perform field normalization processing on the basic information, attribute information, data type, and unit of each device in the acquisition model to obtain a standardized field string; Split the standardized field string into independent fields according to the delimiter and sort them to obtain an ordered field sequence; Use a unified delimiter to sequentially concatenate the ordered field sequence to obtain a concatenated complete string; Apply an encryption hash algorithm to the concatenated complete string to obtain the hash value of the device field.

5. A rapid data acquisition method based on the integration of intelligent question asking and industrial connection technology according to claim 4, characterized in that The steps for combining the fields in the acquisition model into a logical tree structure and performing nested level encoding to obtain the structure path encoding are as follows: Combine the fields in the acquisition model into a logical tree structure, and use the fields of the nodes in the logical tree structure as the encoding result; Obtain the encoding results of all child nodes of the current node in the logical tree structure, and then sort them in lexicographical order to obtain a set of child node encodings; Concatenate all the child node encodings in the set of child node encodings in order into a single string, and apply an encryption hash algorithm to obtain the hash value of the encoding; Take the first 4 digits of the hash value of the encoding as the unique identifier to obtain the aggregated hash value of the child node structure; Concatenate the field of the node with the aggregated hash value of the child node structure. During the concatenation process, if the current node is not the root node, concatenate the parent node path delimiter to form a complete hierarchical path to obtain the structure path encoding.

6. A rapid data acquisition method based on the integration of intelligent question number and industrial connection technology according to claim 5, characterized in that, The steps for using the AI semantic model to transform the device name and device attribute information in the acquisition model into semantic vectors are as follows: Obtain the device name in the acquisition model and perform normalization processing on the device name to obtain a standardized device name; Obtain the device attribute information in the acquisition model, and reorganize the device attribute information in the form of "attribute name: value + unit" to obtain a structured attribute description string; Concatenate the standardized device name with the structured attribute description string as the device complete description text sequence; Perform word segmentation processing on the device complete description text sequence and input it into the pre-trained BERT model, and use the hidden layer vector corresponding to the CLS flag bit output by the BERT model as the semantic representation of the overall text to obtain the semantic encoding vector of the text; Map the semantic encoding vector of the text to the industrial device parameter feature space through a learnable industrial domain adaptation matrix and perform bias adjustment to obtain the domain-adapted semantic vector; Perform non-linear transformation and normalization operations on the domain-adapted semantic vector in sequence to obtain the final semantic vector.

7. A fast data acquisition method based on the integration of intelligent question number and industrial connection technology according to claim 6, characterized in that In step 3 above, based on the structure fingerprint, perform a matching degree analysis on several acquisition models through multi-modal similarity calculation, which specifically includes the following steps: Establish a field mapping relationship for the multi-dimensional fingerprint data according to the field name or business logic for alignment to obtain a multi-modal feature data set with structure alignment; Traverse the set of hash values in the structure-aligned multimodal feature dataset of all acquisition models, count the proportion of the same hash values, and use the proportion of the same hash values as the hash similarity score; Convert the structure path encoding in the structure-aligned multimodal feature dataset into a string sequence; Use the Levenshtein algorithm to calculate the minimum edit distance between the string sequences of all acquisition models and normalize it to obtain the structure path similarity score; Project the semantic vectors in the structure-aligned multimodal feature dataset into a unified semantic space through a pre-trained semantic model; In the unified semantic space, calculate the cosine similarity of the same-name field vectors of all acquisition models to obtain the cosine similarity of all fields; Assign different field weights according to the field importance, and perform a weighted average operation on the cosine similarity of all fields using the field weights to obtain the semantic similarity score; Obtain the required business scenario type and the user's historical operation preferences, and dynamically generate multimodal weights according to the required business scenario type and the user's historical operation preferences to obtain the multimodal weight coefficient; Use the multimodal weight coefficient to perform weighted fusion on the hash similarity score, the structure path similarity score, and the semantic similarity score, and normalize it to obtain the multimodal comprehensive similarity score.

8. A rapid data acquisition method based on the integration of intelligent question number and industrial connection technology according to claim 7, characterized in that, In step 3, screen several acquisition models according to the matching degree analysis result for uniqueness management, and perform an association mapping between the device and the corresponding acquisition model, which specifically includes the following steps: Divide the multimodal comprehensive similarity score into score intervals according to a preset rule and assign corresponding labels to obtain a list of candidate models with classification labels; among them, use greater than or equal to 95% as the high matching interval and label it as "strong reuse candidate"; use 70% - 95% as the gray area and label it as "require manual verification"; use less than 70% as the low matching interval and label it as "independent new creation"; Obtain the set of acquisition models labeled as "strong reuse candidate" in the list of candidate models with classification labels, and perform uniqueness determination. If the same target model only matches one candidate model, directly associate and reuse it; If multiple candidate models match the same target acquisition model, sort them dynamically according to the acquisition model version time, the user's historical selection frequency, and the field integrity; select the acquisition model with the highest priority in the sorting result as the main version, and label the rest as "historical copies"; and extract the fields missing in the main version from the "historical copies" to generate a patch file for manual confirmation to obtain the deduplicated set of main acquisition models; Obtain the set of acquisition models labeled as "require manual verification" in the list of candidate acquisition models with classification labels, and visualize the differences in the similar or conflicting parts to guide the user to correct and update the acquisition models labeled as "require manual verification" to obtain the corrected set of acquisition models; Perform an association mapping between the set of acquisition models labeled as "independent new creation", the deduplicated set of main acquisition models, and the corrected set of acquisition models and the corresponding devices to obtain the deduplicated acquisition models.

9. A fast data acquisition method based on the integration of intelligent question asking and industrial connection technology according to claim 8, characterized in that In step 4, the deduplicated acquisition model is used to accurately distribute the edge gateway configuration through a bidirectional transmission channel. Combining real-time structure fingerprint feedback and matching degree analysis to ensure consistency specifically includes the following steps: Check whether the model fields cover all protocol parameters of the target device for integrity verification; based on the structure identity fingerprint, compare the models already deployed on the target gateway for pre-detection of conflicts; After the integrity verification and pre-detection of conflicts are passed, the verified model is encapsulated into a configuration package with structure identity metadata attached according to the gateway protocol requirements; According to the size and field complexity of the configuration package, select different compression algorithms for compression processing, and select corresponding transmission rules according to the current network environment status to send the configuration package to the edge gateway; During the process of sending the configuration package to the edge gateway, the deployment status is fed back in real time through the bidirectional channel, and it is confirmed whether to enable incremental retransmission according to the deployment status to retransmit the conflicting fields to ensure the integrity of the acquisition model deployment and complete the acquisition model deployment; After the edge gateway completes the deployment of the acquisition model, obtain the actual effective configuration information of the edge gateway, and extract the field-level hash signature and structure path of the gateway configuration at a preset cycle to generate a new fingerprint structure; Compare the difference between the original fingerprint structure at the time of distribution and the new fingerprint structure, and mark the modified fields to obtain real-time structure identity data and difference marks; Obtain the different fields in the real-time structure identity data according to the difference marks, calculate the multi-modal similarity of the different fields for incremental comparison; obtain a real-time consistency analysis report, and perform abnormal configuration repair and policy optimization based on the real-time consistency analysis report.

Images