Optical cable monitoring analysis method and system based on large language model

Abstract

The invention relates to an optical cable monitoring analysis method and system based on a large language model, and the method comprises the steps: obtaining calling information in a calling textbox when a calling instruction sent by a user side is received, and inputting the calling information into a preset calling analysis model; when the calling analysis model receives calling information, semantic recognition processing and identification filling processing are conducted on the calling information, and a calling matching text is obtained; the calling analysis model is used for indexing a preset service program to a matched service node based on the calling matching text, and screening out monitoring data associated with the calling matching text; on the basis of calling the matching text, judging whether the screened monitoring data needs to be analyzed and processed or not; and when analysis processing needs to be executed, acquiring an associated analysis framework based on calling the matching text, sending the monitoring data to the analysis framework for analysis calculation processing, outputting an analysis report and sending the analysis report to a user side. The method has the effect of facilitating targeted calling and analysis of the optical cable monitoring data by the user side.

Description

Technical Field

[0001] The present application relates to the technical field of optical cable detection, and in particular to an optical cable monitoring and analysis method and system based on a large language model. Background Art

[0002] Currently, the purpose of data monitoring of optical cables is to analyze and predict the operating status of optical cables, to timely discover safety hazards and fault locations in optical cable lines, and to improve the reliability and maintenance efficiency of optical cable networks.

[0003] Usually, optical time domain reflectometer is used in combination with environmental data to acquire monitoring data, and the monitoring data is analyzed through a large model to feedback the fault type and fault location to the user. However, the data currently used for large model training is relatively simple, generally only containing monitoring data and geographic location information. Therefore, when the user needs to retrieve the historical monitoring data of the optical cable for analysis, it is necessary to perform multiple data screening steps such as data type, time, location and fault type. The operation is cumbersome and results in a poor user experience, which needs to be improved. Summary of the invention

[0004] In order to facilitate the user end to retrieve and analyze the optical cable monitoring data in a targeted manner and improve the user experience, the present application provides an optical cable monitoring and analysis method and system based on a large language model.

[0005] The above-mentioned invention objective of the present application is achieved through the following technical solutions: A method and system for monitoring and analyzing optical cables based on a large language model, comprising the steps of: When receiving a monitoring data retrieval request from a user terminal, a retrieval text box is sent to the user terminal; When receiving a call instruction sent by the user terminal, obtaining the call information in the call text box and inputting it into a preset call analysis model, the call information including text information; When the call analysis model receives the call information, it performs semantic recognition processing and identifier filling processing on the call information to obtain the call matching text; The retrieval analysis model indexes the matching business nodes from the preset business program based on the retrieval matching text, and filters out the monitoring data associated with the retrieval matching text; Based on the retrieved matching text, determine whether the filtered monitoring data needs to be analyzed and processed; When analysis processing needs to be performed, the associated analysis framework is obtained based on the matching text, the monitoring data is sent to the analysis framework for analysis and calculation processing, and the analysis report is output and sent to the user end.

[0006] By adopting the above technical solution, when the user needs to retrieve historical or current optical cable monitoring data, a monitoring data retrieval request is sent to obtain a retrieval text box for inputting the text content of the monitoring data to be retrieved and sending it to the retrieval analysis model. The retrieval analysis model integrates the text semantic recognition of the retrieval information, the identification information matching of the retrieval information, and the index matching with the business SOP. The text representation of the name, location, time and other information of the monitoring data that the user wants to retrieve is obtained through the semantic analysis of the retrieval information. After the identification filling processing, the name of the monitoring data, the optical cable location number, time and other information can be identified and filled. Filling, identification filling, such as filling of path information and screening filling, so that the obtained retrieved matching text can index the matching business node in the preset business program, and filter out the monitoring data that the user needs to retrieve according to the retrieved matching text, and further judge whether the user has the analysis demand of the monitoring data in the retrieved matching text. If so, the analysis framework is automatically matched for analysis and calculation, and finally the monitoring data is sent to the user; by integrating the large language model of multimodal data, the recognition and conversion of monitoring data and the analysis and matching of the user's willingness to retrieve monitoring data are realized. For the user end, the operation is simple and the user experience is better.

[0007] Optionally, when the call analysis model receives the call information, it performs semantic recognition processing and identifier filling processing on the call information to obtain the call matching text; including: The retrieval analysis model outputs semantic text corresponding to the retrieval information based on semantic recognition processing, wherein the semantic recognition includes context perception and semantic reasoning; Based on context perception, a preset field is extracted from the semantic text, and the preset field includes time information, optical cable location number information and data name information of the monitoring data to be retrieved, and the time information and optical cable location number information in the same preset field are the same; The preset field is subjected to identification filling processing, which includes retrieving identification information matching the preset field from a preset database and filling it into the preset field of the semantic text to obtain retrieved matching text, wherein the identification information includes path information.

[0008] By adopting the above technical solution, through context perception and semantic reasoning, the retrieval analysis model outputs the received retrieval information in a unified text format and text expression logic as a semantic text, accurately knows the content that the user end wants to obtain, and extracts the preset field from the semantic text. The preset field represents the information of the monitoring data that the user end wants to retrieve, including data name, optical cable location number, data collection time and other information, and the time information and location information in the same preset field are the same, which is convenient for subsequent retrieval of the business node index of the matching text in the preset business program. For example, through context perception, a preset field in the semantic text is obtained, which is the monitoring data D of the optical cable with A time, B location and number C. Then, the preset field is filled with identification information, and the identification information includes A time, B location, optical cable number C and path information of the monitoring data D; so as to obtain the retrieval matching text for matching the business node.

[0009] Optionally, the retrieved analysis model indexes a matching business node from a preset business program based on the retrieved matching text, and filters out monitoring data associated with the retrieved matching text, including: Retrieving the analysis model to identify and retrieve identification information of different preset fields in the matching text; and indexing the preset fields one by one; Indexing to the service node in the service program based on the time information and the optical cable location number information in the preset field; Based on the data name information in the preset field, the monitoring data associated with the retrieved matching text is filtered out from all the monitoring data of the business node.

[0010] By adopting the above technical solution, the retrieval analysis model first identifies all the preset fields in the retrieval matching text, indexes the monitoring data with different time information and location information separately, improves the indexing efficiency and accuracy, and indexes the service node of the target optical cable based on the time and optical cable location number. At this time, all the monitoring data of the service node are indexed, and then the monitoring data that the user wants to retrieve is filtered out according to the data name information.

[0011] Optionally, the retrieval analysis model outputs a semantic text corresponding to the retrieval information based on semantic recognition processing, wherein the semantic recognition includes context awareness and semantic reasoning; including: When the retrieval analysis model receives the retrieval information, it identifies the text features of the current retrieval information based on the pre-trained context-aware relationship and semantic reasoning decision, and determines the text format information of the current retrieval information; Get the text conversion template that matches the currently determined text format information; Based on the text conversion template, the retrieved information is converted into semantic text.

[0012] By adopting the above technical solution, since the retrieved information is text, the user end may express the monitoring data to be retrieved through different text formats. Therefore, a knowledge graph is constructed in advance and a large amount of corpus is fed to the retrieved analysis model. The context-aware relationship and the decision logic of semantic reasoning are trained through a large amount of retrieved information to obtain a large amount of customized text format information. The text format information of the current retrieved information is further obtained through the recognition of text features. The retrieved information is converted into a semantic text in a unified text format through a pre-stored text conversion template for each text format information. For example: the user end inputs and retrieves all monitoring data of the optical cable number C at A time and B position, but excludes D data in all monitoring data. It is known that the monitoring data of the optical cable number C includes DEFG, which can be uniformly converted into: EFG data of the optical cable number C at A time and B position through a text conversion template. The unified semantic text facilitates the efficiency of subsequent indexing.

[0013] Optionally, the identifying of text features of the currently retrieved information and determining the text format information of the currently retrieved information include: the text features include semantic types represented by different fields in the retrieved information, and semantic structural features formed after sorting different semantic types in the retrieved information.

[0014] By adopting the above technical solution, the semantic structure characteristics of various semantic types in the text are obtained by analyzing the text of the retrieved information, wherein the semantic types include the semantics of the execution means such as retrieval, acquisition, spot check, screening, etc., and also include the names and nouns representing certain information and various common pronouns and conjunctions, thereby realizing the determination of the text format information of the retrieved information.

[0015] Optionally, the step of determining whether the filtered monitoring data needs to be analyzed and processed based on the retrieved matching text includes: Traverse the retrieved matching text, identify and determine whether the retrieved matching text contains the analysis requirement field based on the context, and the analysis field contains the specific analysis method; If the analysis requirement field is included, it is determined that the currently retrieved matching text needs to be analyzed and processed; If the analysis requirement field is not included, it is determined that the current retrieved matching text does not need to be analyzed and processed, and the filtered monitoring data is sent to the user end.

[0016] By adopting the above technical solution, by identifying the fields in the retrieved matching text, including identifying the fields representing the execution analysis and the name of the analysis algorithm, it is determined whether the current monitoring data needs to be analyzed. The identification method is more flexible. If the analysis requirement field is not identified, the retrieved matching text is directly sent to the user end.

[0017] Optionally, when the analysis process needs to be performed, the associated analysis framework is obtained based on the matching text, the monitoring data is sent to the analysis framework for analysis and calculation, and the analysis report is output and sent to the user end, including: Based on the required analysis fields, the associated analysis framework is obtained from the preset algorithm database, and the analysis framework includes the analysis algorithm and the report template; The monitoring data is calculated through the associated analysis algorithm, and the calculation results are filled in the corresponding position of the report template to generate an analysis report and send it to the user end.

[0018] By adopting the above technical solution, based on the required analysis field, the associated analysis algorithm is obtained and the monitoring data is substituted into the analysis algorithm for automatic calculation. The monitoring data and the calculation results are mapped based on the calculation result mapping relationship of the analysis framework, filled in the corresponding position in the report template and sent to the user end, so as to facilitate the user end to browse the data.

[0019] The second object of the invention is achieved by the following technical solutions: An optical cable monitoring and analysis system based on a large language model, comprising: The retrieval request module is used to send a retrieval text box to the user end when receiving a monitoring data retrieval request sent by the user end; An information acquisition module, for acquiring the retrieval information in the retrieval text box and inputting it into a preset retrieval analysis model when receiving a retrieval instruction sent by the user end, wherein the retrieval information includes text information; A semantic recognition module, which is used to perform semantic recognition processing and identification filling processing on the retrieved information when the retrieved analysis model receives the retrieved information, so as to obtain the retrieved matching text; A matching module, used for the retrieved analysis model to index the matching business node from the preset business program based on the retrieved matching text, and filter out the monitoring data associated with the retrieved matching text; An analysis and judgment module is used to judge whether the filtered monitoring data needs to be analyzed and processed based on the retrieved matching text; The analysis sending module is used to obtain the associated analysis framework based on the matching text when analysis processing needs to be performed, send the monitoring data to the analysis framework for analysis and calculation processing, output the analysis report and send it to the user end.

[0020] The third objective of the present application is achieved through the following technical solutions: A computer device comprises a memory, a processor and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, the steps of the optical cable monitoring and analysis method based on a large language model are implemented.

[0021] The fourth objective of the present application is achieved through the following technical solutions: A computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of the optical cable monitoring and analysis method based on a large language model are implemented.

[0022] In summary, the present application includes at least one of the following beneficial technical effects: 1. After the identification filling process, the name, optical cable location number, time and other information of the monitoring data are identified and filled. The identification filling, for example, the filling of path information and the screening filling, can be used to index the matching service node in the preset service program, and filter out the monitoring data that the user needs to retrieve according to the retrieved matching text, and further judge whether the user has the need to analyze the monitoring data in the retrieved matching text. If so, the analysis framework is automatically matched for analysis and calculation, and finally the monitoring data is sent to the user. By integrating the large language model of multimodal data, the recognition and conversion of monitoring data and the analysis and matching of the user's willingness to retrieve monitoring data are realized. For the user, the operation is simple and the user experience is better. 2. The retrieval analysis model first identifies all preset fields in the retrieval matching text, and indexes the monitoring data with different time information and location information separately to improve the indexing efficiency and accuracy. The service node of the target optical cable is indexed based on the time and optical cable location number. At this time, all the monitoring data of the service node is indexed. Then, the monitoring data that the user wants to retrieve is filtered out according to the data name information; 3. Convert the retrieved information into a semantic text in a unified text format. For example, the user inputs and retrieves all monitoring data of the optical cable numbered C at time A and position B, but excludes data D from all monitoring data. It is known that the monitoring data of the optical cable numbered C includes DEFG. Then, the text conversion template can be used to uniformly convert the data into: retrieve the data of the optical cable numbered C at time A and position B, and the unified semantic text facilitates the efficiency of subsequent indexing. 4. By analyzing the text of the retrieved information, the semantic structural features composed of various semantic types in the text are obtained, among which the semantic types include the semantics of the execution means such as retrieval, acquisition, spot check, and screening, as well as the names and nouns representing certain information and various common pronouns and conjunctions, so as to realize the judgment of the text format information of the retrieved information. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 This is a flowchart of an implementation of an optical cable monitoring and analysis method based on a large language model in the present application; Figure 2This is a flowchart for implementing step S30 in an embodiment of an optical cable monitoring and analysis method based on a large language model of the present application; Figure 3 This is a flowchart for implementing step S31 in an embodiment of an optical cable monitoring and analysis method based on a large language model of the present application; Figure 4 This is a flowchart for implementing step S40 in an embodiment of an optical cable monitoring and analysis method based on a large language model of the present application; Figure 5 It is a schematic diagram of a computer device of the present application. DETAILED DESCRIPTION

[0024] The following is combined with Figure 1-5 This application is described in further detail.

[0025] In the following embodiments, Figure 1 As shown, the present application discloses an optical cable monitoring and analysis method based on a large language model, which specifically includes the following steps: S10: When receiving a monitoring data retrieval request from the user terminal, sending a retrieval text box to the user terminal; In this embodiment, the user end refers to a PC or mobile terminal software used by optical cable data retrieval personnel. The monitoring data of the optical cable includes but is not limited to optical cable deformation data, optical cable temperature data, temperature data, vibration and intrusion signal data, bit error rate, optical signal-to-noise ratio, and environmental data of the optical cable burial site.

[0026] The user terminal issues a data retrieval request to obtain a retrieval text box for inputting text on the user terminal. Optionally, it also includes sending a picture input text box or a voice input box.

[0027] Specifically, the optical cable data retrieval personnel send an optical cable monitoring data retrieval request instruction through the user terminal to obtain a retrieval text box on the user terminal for inputting the text of the monitoring data to be retrieved.

[0028] S20: when receiving the call instruction sent by the user terminal, obtaining the call information in the call text box and inputting it into the preset call analysis model, the call information includes text information; In this embodiment, the retrieval information refers to the text of the monitoring data to be retrieved, and the retrieval instruction refers to the instruction issued by the user end to trigger the retrieval information to be sent to the retrieval analysis model.

[0029] S30: When the retrieval analysis model receives the retrieval information, it performs semantic recognition processing and identifier filling processing on the retrieval information to obtain the retrieval matching text; In this embodiment, the retrieval analysis model is a large language model trained using a neural network, which integrates the functions of text semantic recognition, text format conversion, text filling, and matching text with business nodes of preset business programs. The semantic recognition processing includes format recognition of the text of the retrieved information and conversion of the text format and sentence structure. The purpose is to unify the text structure of the retrieved information to facilitate the subsequent indexing of business nodes of the text in the retrieved information. The identification filling processing is to fill the path index information of the monitoring data storage location in the text of the retrieved information.

[0030] Specifically, when the retrieval analysis model receives the text of the retrieval information, it performs semantic recognition processing on the text to convert the text format and sentence structure of the retrieval information, and fills in the path index information of the monitoring data at the location in the text of the retrieval information that expresses the monitoring data to be retrieved. The retrieved matching text is the text that completes the text format conversion and is filled with the index path information of the monitoring data.

[0031] S40: the retrieved analysis model indexes a matching business node from a preset business program based on the retrieved matching text, and filters out monitoring data associated with the retrieved matching text; In this embodiment, the preset business procedure is a storage path SOP process for different monitoring data customized according to business needs, which is equivalent to a storage path tree diagram. The storage locations of different monitoring data can be indexed through the preset business procedure, and each storage location is a business node.

[0032] Specifically, the retrieval analysis model indexes from the preset business GOP process to the business node where the monitoring data to be retrieved by the retrieval matching text exists based on the semantics in the retrieval matching text and the filled path index information, and again filters out the related monitoring data based on the content of the retrieval matching text.

[0033] S50: Based on the retrieved matching text, determine whether the screened monitoring data needs to be analyzed and processed; In this embodiment, the analysis and processing includes, but is not limited to, trend analysis of indicator data, predictive analysis of data trends, distribution analysis of abnormal indicator data, and emergency plan analysis of abnormal indicator data.

[0034] Specifically, character semantic recognition is performed based on the retrieved matching text content, so as to determine whether the monitoring data screened this time needs to be analyzed and processed.

[0035] S60: When analysis processing needs to be performed, the associated analysis framework is obtained based on the matching text, the monitoring data is sent to the analysis framework for analysis and calculation processing, and an analysis report is output and sent to the user end.

[0036] In this embodiment, the analysis framework includes calculation logic steps arranged in a preset order and a corresponding analysis report template. Each logic step is provided with an algorithm formula, and each logic step is used to identify and obtain corresponding monitoring data and fill in the calculation position corresponding to the algorithm formula. After the calculation is completed, the generated data is mapped to the report template.

[0037] Specifically, when text representing the need for monitoring data analysis on the user side is identified from the retrieved matching text, the monitoring data is sent to the algorithm formula calculation position of the corresponding logical step in the analysis framework based on the analysis framework associated with the identified text. After step-by-step calculation according to the arrangement of the algorithm formula, the generated calculation results are input into the corresponding analysis report template to generate an analysis report, and then sent to the user side.

[0038] In one embodiment, referring to Figure 2 , step S30 comprises the steps of: S31: the retrieval analysis model outputs a semantic text corresponding to the retrieval information based on semantic recognition processing, wherein the semantic recognition includes context perception and semantic reasoning; S32: extracting a preset field from the semantic text based on context perception, the preset field including time information, optical cable location number information and data name information of the monitoring data to be retrieved, and the time information and optical cable location number information in the same preset field are the same; S33: performing identification filling processing on the preset field, the identification filling processing includes retrieving identification information matching the preset field from a preset database, and filling it into the preset field of the semantic text to obtain the retrieved matching text, wherein the identification information includes path information.

[0039] In this embodiment, semantic reasoning aims to extract implicit logical relationships, intentions, and knowledge from text, and obtain new conclusions through reasoning. Its essence is to establish a connection between language symbols and real-world knowledge, which mainly relies on knowledge representation, logical reasoning, and distributed semantics. Context awareness means that the system dynamically adjusts its understanding and response based on the current conversation, environment, or user historical behavior. The key lies in capturing and utilizing multi-dimensional context information, including local context, global context, and dynamic context, which is achieved by using attention mechanisms, memory networks, and state tracking.

[0040] The preset fields are distinguished by the position of the optical cable number. For example, if the optical cable at position A is numbered B, the monitoring data of the same preset field are all the optical cables at position A is numbered B, and the number can be one or more.

[0041] Path information is path index information, which is used to index the business node where the monitoring data is stored in the business SOP. The preset field in the semantic text is ABCDE, and its corresponding path information is 1 / 2 / 3. After the identification filling process, the ABCDE text in the matching text will be displayed as ABCDE (1 / 2 / 3).

[0042] In one embodiment, referring to Figure 3 , step S31 comprises the steps of: S311: When the retrieval analysis model receives the retrieval information, based on the pre-trained context-aware relationship and semantic reasoning decision, it identifies the text features of the current retrieval information and determines the text format information of the current retrieval information; S312: Obtaining a text conversion template that matches the currently determined text format information; S313: Based on the text conversion template, convert the retrieved information into semantic text.

[0043] In this embodiment, the text features include semantic types represented by different fields in the retrieved information, and semantic structural features formed after the different semantic types are sorted in the retrieved information; Specifically, the conversion of text format is combined with context perception and semantic reasoning decision-making to realize the disassembly of text format at the semantic level and the conversion of text sorting. For example, the information retrieved is to retrieve the monitoring data of all types of optical cables at position A. After semantic recognition and database comparison, the analysis shows that there are three types of optical cables, BCD, at position A. The semantic text is converted into the monitoring data of optical cable model B at position A, the monitoring data of optical cable model C at position A, and the monitoring data of optical cable model D at position A, that is, the disassembly of three preset fields is formed, and the text format is in the format of location-model-monitoring data, thereby realizing the disassembly of the preset fields.

[0044] In addition, the conversion process needs to be combined with the semantics of the identification filter words, including but not limited to the filter words such as acquisition, screening and exclusion.

[0045] In one embodiment, referring to Figure 4 , step S40 comprises the steps of: S41: Retrieving the analysis model to identify and retrieve identification information of different preset fields in the matching text; and indexing the preset fields one by one; S42: indexing to a service node in a service program based on the time information and the optical cable position number information in the preset field; S43: Based on the data name information in the preset field, the monitoring data associated with the retrieved matching text is screened out from all the monitoring data of the business node.

[0046] In this embodiment, each service node stores all monitoring data corresponding to the optical cable position number. The specific monitoring data to be retrieved needs to be further screened and identified through the data name information in the preset field to obtain all, part or a single monitoring data.

[0047] In one embodiment, step S50 includes the steps of: S51: traverse the retrieved matching text, identify and determine whether the retrieved matching text contains an analysis requirement field based on the context, and the analysis field contains a specific analysis method; S52: If the analysis requirement field is included, it is determined that the currently retrieved matching text needs to be analyzed and processed; S53: If the analysis requirement field is not included, it is determined that the currently retrieved matching text does not need to be analyzed and processed, and the filtered monitoring data is sent to the user end.

[0048] In this embodiment, the analysis requirement field includes not only the analysis method, but also the context part of the analysis method, such as the word "not" for not analyzing. The known analysis methods are pre-stored in the database, and the analysis requirement field is semantically identified and matched with the known analysis methods in the database to determine whether the user terminal needs to perform analysis processing.

[0049] In one embodiment, step S60 includes the steps of: S61: based on the required analysis field, obtaining a related analysis framework from a preset algorithm database, the analysis framework including an analysis algorithm and a report template; S62: Calculate the monitoring data through the associated analysis algorithm, fill the calculation results into the corresponding position of the report template, generate an analysis report and send it to the user end.

[0050] In this embodiment, the analysis algorithm is the algorithm formula in each logical step.

[0051] It should be understood that the size of the serial numbers of the steps in the above embodiments does not mean the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present application.

[0052] In one embodiment, an optical cable monitoring and analysis system based on a large language model is provided, and the optical cable monitoring and analysis system based on a large language model corresponds one-to-one to an optical cable monitoring and analysis method based on a large language model in the above embodiment. An optical cable monitoring and analysis system based on a large language model includes: The retrieval request module is used to send a retrieval text box to the user end when receiving a monitoring data retrieval request sent by the user end; An information acquisition module, for acquiring the retrieval information in the retrieval text box and inputting it into a preset retrieval analysis model when receiving a retrieval instruction sent by the user end, wherein the retrieval information includes text information; A semantic recognition module, which is used to perform semantic recognition processing and identification filling processing on the retrieved information when the retrieved analysis model receives the retrieved information, so as to obtain the retrieved matching text; A matching module, used for the retrieved analysis model to index the matching business node from the preset business program based on the retrieved matching text, and filter out the monitoring data associated with the retrieved matching text; An analysis and judgment module is used to judge whether the filtered monitoring data needs to be analyzed and processed based on the retrieved matching text; The analysis sending module is used to obtain the associated analysis framework based on the matching text when analysis processing needs to be performed, send the monitoring data to the analysis framework for analysis and calculation processing, output the analysis report and send it to the user end.

[0053] Optionally, the semantic recognition module includes: A semantic text submodule, which is used for the retrieval analysis model to process and output semantic text corresponding to the retrieved information based on semantic recognition, wherein the semantic recognition includes context perception and semantic reasoning; The preset field submodule is used to extract the preset field from the semantic text based on context perception. The preset field includes the time information, optical cable position number information and data name information of the monitoring data to be retrieved, and the time information and optical cable position number information in the same preset field are the same; The filling submodule is used to perform identification filling processing on the preset field. The identification filling processing includes retrieving identification information matching the preset field from the preset database and filling it into the preset field of the semantic text to obtain the retrieved matching text, and the identification information includes path information.

[0054] Optionally, matching modules include: The index identification submodule is used to retrieve the identification information of different preset fields in the matching text by using the analysis model to identify and retrieve; and to index the preset fields one by one; An indexing submodule, used to index to a service node in a service program based on time information and optical cable position number information in a preset field; The index filtering submodule is used to filter out the monitoring data associated with the retrieved matching text from all the monitoring data of the business node based on the data name information in the preset field.

[0055] Optional, semantic text submodules include: A text feature recognition unit is used to recognize text features of the current retrieved information and determine the text format information of the current retrieved information based on pre-trained context-aware relationships and semantic reasoning decisions when the retrieved analysis model receives the retrieved information; the text features include semantic types represented by different fields in the retrieved information, and semantic structural features formed after different semantic types are sorted in the retrieved information; A text format recognition unit, used to obtain a text conversion template that matches the currently determined text format information; The text conversion unit converts the retrieved information into semantic text based on the text conversion template.

[0056] For the specific definition of an optical cable monitoring and analysis system based on a large language model, please refer to the definition of an optical cable monitoring and analysis method based on a large language model in the above text, which will not be repeated here.

[0057] Optionally, the analysis and judgment module includes: The analysis and identification submodule is used to traverse the retrieved matching text, identify and determine whether the retrieved matching text contains the analysis requirement field based on the context, and the analysis field contains the specific analysis method; A first analysis execution module, used to determine that the currently retrieved matching text needs to be analyzed and processed if an analysis requirement field is included; The second analysis execution module is used to determine that the currently retrieved matching text does not need to be analyzed and processed if the analysis requirement field is not included, and then send the filtered monitoring data to the user end.

[0058] Optionally, the analysis and sending module includes: The framework acquisition submodule is used to obtain the associated analysis framework from the preset algorithm database based on the required analysis field. The analysis framework includes the analysis algorithm and the report template; The calculation submodule is used to calculate the monitoring data through the associated analysis algorithm, fill the calculation results into the corresponding position of the report template, generate an analysis report and send it to the user end.

[0059] Each module in the above optical cable monitoring and analysis system based on a large language model can be implemented in whole or in part by software, hardware, or a combination thereof. Each module can be embedded in or independent of a processor in a computer device in the form of hardware, or can be stored in a memory in a computer device in the form of software, so that the processor can call and execute operations corresponding to each module.

[0060] In one embodiment, a computer device is provided. The computer device may be a server, and its internal structure diagram may be as follows: Figure 5As shown. The computer device includes a processor, a memory, a network interface and a database connected through a system bus. Among them, the processor of the computer device is used to provide computing and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program and a database. The internal memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The network interface of the computer device is used to communicate with an external terminal through a network connection. When the computer program is executed by the processor, a method for optical cable monitoring and analysis based on a large language model is implemented.

[0061] In one embodiment, a computer device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, a method for monitoring and analyzing an optical cable based on a large language model is implemented.

[0062] In one embodiment, a computer-readable storage medium is provided, on which a computer program is stored. When the computer program is executed by a processor, an optical cable monitoring and analysis method based on a large language model is implemented.

[0063] Those of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing the relevant hardware through a computer program, and the computer program can be stored in a non-volatile computer-readable storage medium. When the computer program is executed, it can include the processes of the embodiments of the above-mentioned methods. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0064] Those skilled in the art can clearly understand that for the convenience and simplicity of description, only the division of the above-mentioned functional units and modules is used as an example. In actual applications, the above-mentioned functions can be distributed and completed by different functional units and modules as needed, that is, the internal structure of the device can be divided into different functional units or modules to complete all or part of the functions described above.

[0065] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Although the present application has been described in detail with reference to the aforementioned embodiments, a person skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. An optical cable monitoring and analysis method based on a large language model, characterized in that: When receiving a monitoring data retrieval request from a user terminal, a retrieval text box is sent to the user terminal; When receiving a call instruction sent by the user terminal, obtaining the call information in the call text box and inputting it into a preset call analysis model, the call information including text information; When the call analysis model receives the call information, it performs semantic recognition processing and identifier filling processing on the call information to obtain the call matching text; The retrieval analysis model retrieves the matching text, indexes the matching business node from the preset business program, and filters out the monitoring data associated with the retrieved matching text; Based on the retrieved matching text, determine whether the filtered monitoring data needs to be analyzed and processed; When analysis processing needs to be performed, the associated analysis framework is obtained based on the matching text, the monitoring data is sent to the analysis framework for analysis and calculation processing, and the analysis report is output and sent to the user end.

2. The optical cable monitoring and analysis method based on a large language model according to claim 1 is characterized in that: When the call analysis model receives the call information, it performs semantic recognition processing and identifier filling processing on the call information to obtain the call matching text; including: The retrieval analysis model outputs semantic text corresponding to the retrieval information based on semantic recognition processing, wherein the semantic recognition includes context perception and semantic reasoning; Based on context perception, a preset field is extracted from the semantic text, and the preset field includes time information, optical cable location number information and data name information of the monitoring data to be retrieved, and the time information and optical cable location number information in the same preset field are the same; The preset field is subjected to identification filling processing, which includes retrieving identification information matching the preset field from a preset database and filling it into the preset field of the semantic text to obtain retrieved matching text, wherein the identification information includes path information.

3. The optical cable monitoring and analysis method based on a large language model according to claim 1 is characterized in that: The retrieval analysis model retrieves matching texts, indexes matching business nodes from preset business programs, and filters out monitoring data associated with the retrieved matching texts, including: Retrieving the analysis model to identify and retrieve identification information of different preset fields in the matching text; and indexing the preset fields one by one; Indexing to the service node in the service program based on the time information and the optical cable location number information in the preset field; Based on the data name information in the preset field, the monitoring data associated with the retrieved matching text is filtered out from all the monitoring data of the business node.

4. The optical cable monitoring and analysis method based on a large language model according to claim 2 is characterized in that: The retrieval analysis model outputs semantic text corresponding to the retrieval information based on semantic recognition processing, wherein the semantic recognition includes context perception and semantic reasoning; including: When the retrieval analysis model receives the retrieval information, it identifies the text features of the current retrieval information based on the pre-trained context-aware relationship and semantic reasoning decision, and determines the text format information of the current retrieval information; Get the text conversion template that matches the currently determined text format information; Based on the text conversion template, the retrieved information is converted into semantic text.

5. The optical cable monitoring and analysis method based on a large language model according to claim 4 is characterized in that: The identifying of text features of the currently retrieved information and determining the text format information of the currently retrieved information include: the text features include semantic types represented by different fields in the retrieved information, and semantic structure features formed after different semantic types are sorted in the retrieved information.

6. The optical cable monitoring and analysis method based on a large language model according to claim 1 is characterized in that: The method of determining whether the screened monitoring data needs to be analyzed and processed based on the retrieved matching text includes: Traverse the retrieved matching text, identify and determine whether the retrieved matching text contains the analysis requirement field based on the context, and the analysis field contains the specific analysis method; If the analysis requirement field is included, it is determined that the currently retrieved matching text needs to be analyzed and processed; If the analysis requirement field is not included, it is determined that the current retrieved matching text does not need to be analyzed and processed, and the filtered monitoring data is sent to the user end.

7. The optical cable monitoring and analysis method based on a large language model according to claim 6 is characterized in that: When the analysis process needs to be performed, the associated analysis framework is obtained based on the matching text, the monitoring data is sent to the analysis framework for analysis and calculation, and the analysis report is output and sent to the user end, including: Based on the required analysis fields, the associated analysis framework is obtained from the preset algorithm database, and the analysis framework includes the analysis algorithm and the report template; The monitoring data is calculated through the associated analysis algorithm, and the calculation results are filled in the corresponding position of the report template to generate an analysis report and send it to the user end.

8. An optical cable monitoring and analysis system based on a large language model, characterized in that: The retrieval request module is used to send a retrieval text box to the user end when receiving a monitoring data retrieval request sent by the user end; An information acquisition module, for acquiring the retrieval information in the retrieval text box and inputting it into a preset retrieval analysis model when receiving a retrieval instruction sent by the user end, wherein the retrieval information includes text information; A semantic recognition module, which is used to perform semantic recognition processing and identification filling processing on the retrieved information when the retrieved analysis model receives the retrieved information, so as to obtain the retrieved matching text; A matching module, used for the retrieved analysis model to index the matching business node from the preset business program based on the retrieved matching text, and filter out the monitoring data associated with the retrieved matching text; An analysis and judgment module is used to judge whether the filtered monitoring data needs to be analyzed and processed based on the retrieved matching text; The analysis sending module is used to obtain the associated analysis framework based on the matching text when analysis processing needs to be performed, send the monitoring data to the analysis framework for analysis and calculation processing, output the analysis report and send it to the user end.

9. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the optical cable monitoring and analysis method based on a large language model as described in any one of claims 1 to 7 are implemented.

10. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the steps of the optical cable monitoring and analysis method based on a large language model as claimed in any one of claims 1 to 7 are implemented.

Images