A dedicated line comprises a data management system, a method, and a storage medium.

Abstract

This invention provides a dedicated line routing data management system, method, and storage medium, belonging to the field of routing management. It includes a data access layer, an intelligent core layer, a support layer, and an application interaction layer. The data access layer collects multiple raw dedicated line routing data sets and performs standardization processing on each set to obtain standardized dedicated line routing data. This invention solves the problems of automatically converting unstructured raw information into high-quality network resource data, automatically implementing a complete closed-loop handling process after route construction failure (from root cause diagnosis, task dispatch, repair execution to result verification), and performing intelligent reasoning and decision-making based on domain knowledge in complex and uncertain scenarios, driving automatic collaborative operation of multiple heterogeneous systems.

Description

Technical Field

[0001] This invention mainly relates to the field of routing management technology, specifically to a dedicated line routing data management system, method, and storage medium. Background Technology

[0002] With the rapid development of the digital economy, government and enterprise customers are increasingly demanding high-quality, highly reliable, and transparently controllable leased line services. To meet customers' needs for visualized monitoring and intelligent operation and maintenance of end-to-end leased line service quality, telecom operators urgently need to build a complete physical routing view that can connect the customer's access point to the core network, spanning multiple vendors' equipment and multi-layered network architectures. This technical process is commonly referred to in the industry as "leased line routing splicing." Essentially, it uses algorithms and systems to intelligently associate and reconstruct the configuration data of devices, ports, and links scattered across resource management systems, scheduling systems, and network element management systems, thereby logically splicing together an end-to-end routing topology that reflects the actual physical connection relationships.

[0003] Currently, there are three main technical problems in the field of dedicated line routing data management for government and enterprise businesses: (1) The lack of intelligent data source governance leads to a high reliance on manual labor and low efficiency in the initial stages of the process. Existing technologies cannot effectively understand and process multi-source, heterogeneous, and unstructured raw information (such as free text descriptions and non-standard tables) from the construction and planning stages. At the same time, accurately converting this information into high-quality structured network resource data required by subsequent IT systems relies entirely on the manual interpretation, translation, and input of domain experts. This introduces the risk of human error and severely restricts overall efficiency.

[0004] (2) The anomaly handling process is in an open-loop state, lacking end-to-end automated closed-loop capability from diagnosis to repair. When the routing splicing process fails due to data problems or system conflicts, existing technologies can usually only output preliminary failure indicators. From accurate root cause location of the failure point, to automatic cross-system responsibility determination and task assignment, and then to the execution of repair actions and result verification, this complete handling chain heavily relies on manual offline communication, coordination and operation, making the system itself unable to drive closed-loop problem resolution.

[0005] The existing solutions suffer from insufficient inter-system collaboration and intelligent decision-making for complex scenarios, making them unable to adaptively handle non-standard scenarios. They heavily rely on pre-defined, rigid business rules for data verification and path calculation, lacking a deep understanding of network domain knowledge and flexible reasoning capabilities, such as business logic, device semantics, and transmission network topology constraints. For new device models, novel network configurations, or complex anomaly scenarios not explicitly defined in the rule base, the system exhibits poor adaptability and generalization capabilities. Furthermore, collaborative operations across multiple independent heterogeneous systems, such as scheduling systems, resource centers, splicing systems, and network management systems, require manual switching between different interfaces, hindering intelligent decision-making and automatic scheduling based on a unified strategy. Summary of the Invention

[0006] The technical problem to be solved by the present invention is to provide a dedicated line routing data management system, method and storage medium to address the shortcomings of the prior art.

[0007] The technical solution of this invention to solve the above-mentioned technical problems is as follows: A dedicated line traffic data management system, comprising a data access layer, an intelligent core layer, a support layer, and an application interaction layer. The data access layer is used to collect multiple raw dedicated line traffic data, and to standardize each raw dedicated line traffic data to obtain standardized dedicated line traffic data corresponding to each raw dedicated line traffic data. The intelligent core layer is used to preprocess each of the standardized dedicated line traffic data to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data. The support layer is used to obtain a route optimization database, a route repair database, and multiple raw device parameters from a preset domain knowledge base; The intelligent core layer is also used to identify and analyze all the preprocessed dedicated line route texts based on all the original equipment parameters to obtain target resource object information; The application interaction layer is used to import business intent commands; The intelligent core layer is also used to perform fault analysis on the target resource object information according to the business intent instruction, the route optimization database and the route repair database, and generate route repair results based on the analysis results.

[0008] Another technical solution of the present invention to solve the above-mentioned technical problems is as follows: A dedicated line route data management method, comprising the following steps: Multiple raw dedicated line routing data are collected, and each raw dedicated line routing data is standardized to obtain standardized dedicated line routing data corresponding to each raw dedicated line routing data. Each of the standardized dedicated line traffic data is preprocessed to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data. Obtain route optimization database, route repair database, and multiple raw device parameters from the preset domain knowledge base; Based on all the original equipment parameters, all the preprocessed dedicated line route texts are identified and analyzed to obtain target resource object information; Import business intent instructions, perform fault analysis on the target resource object information based on the business intent instructions, the route optimization database, and the route repair database, and generate route repair results based on the analysis results.

[0009] The beneficial effects of this invention are as follows: Standardized dedicated line routing data is obtained through the standardization of original dedicated line routing data; preprocessing of the standardized dedicated line routing data yields preprocessed dedicated line routing text; routing optimization database, routing repair database, and original equipment parameters are obtained from a preset domain knowledge base; target resource object information is obtained through identification and analysis of the preprocessed dedicated line routing text based on the original equipment parameters; fault analysis is performed on the target resource object information based on business intent instructions, the routing optimization database, and the routing repair database; and routing repair results are generated based on the analysis results. This solves the problems of automatically converting unstructured raw information into high-quality network resource data, automatically implementing a complete closed-loop handling process from root cause diagnosis, task assignment, repair execution to result verification after route construction failure, and performing intelligent reasoning and decision-making based on domain knowledge in complex and uncertain scenarios, driving automatic collaborative operation of multiple heterogeneous systems. It eliminates the high dependence on manual input, improves the intelligence level and accuracy of data source processing, replaces the current inefficient manual offline supervision mode, significantly shortens the problem-solving cycle, and improves the adaptability, reliability, and overall intelligence level of the route construction process. Attached Figure Description

[0010] Figure 1 A schematic diagram of the modules of a dedicated line route data management system provided in an embodiment of the present invention; Figure 2 This is a block diagram of the overall architecture of a dedicated line traffic data management system provided in an embodiment of the present invention; Figure 3 A flowchart illustrating the technical solution of a dedicated line traffic data management system provided in an embodiment of the present invention; Figure 4 This is a flowchart illustrating a dedicated line traffic data management method according to an embodiment of the present invention. Detailed Implementation

[0011] The principles and features of the present invention are described below with reference to the accompanying drawings. The examples given are only for explaining the present invention and are not intended to limit the scope of the present invention.

[0012] Figure 1 This is a block diagram of a dedicated line traffic data management system provided in an embodiment of the present invention.

[0013] like Figure 1 As shown, a dedicated line traffic data management system includes a data access layer, an intelligent core layer, a support layer, and an application interaction layer. The data access layer is used to collect multiple raw dedicated line traffic data, and to standardize each raw dedicated line traffic data to obtain standardized dedicated line traffic data corresponding to each raw dedicated line traffic data. The intelligent core layer is used to preprocess each of the standardized dedicated line traffic data to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data. The support layer is used to obtain a route optimization database, a route repair database, and multiple raw device parameters from a preset domain knowledge base; The intelligent core layer is also used to identify and analyze all the preprocessed dedicated line route texts based on all the original equipment parameters to obtain target resource object information; The application interaction layer is used to import business intent commands; The intelligent core layer is also used to perform fault analysis on the target resource object information according to the business intent instruction, the route optimization database and the route repair database, and generate route repair results based on the analysis results.

[0014] It should be understood that the system adopts a layered architecture, which includes a data access layer, an intelligent core layer, and an application interaction layer from bottom to top, and is supported throughout by a domain knowledge base.

[0015] Specifically, the data access layer serves as the unified input port for the system, connecting the scheduling work order system, network resource center, various professional network management systems, and construction document library. Its function is to collect original work orders, resource configurations, alarms, and unstructured construction reports (i.e., original dedicated line routing data) in real time or on a timed basis, standardize their formats, and encapsulate them into a unified data event stream for output.

[0016] It should be understood that the application interaction layer serves as the human-computer interaction interface, providing personnel with panoramic visual monitoring, natural language intent input, and process intervention management entry points.

[0017] It should be understood that the support layer stores structured knowledge such as network device models, port mapping rules, business logic, time slot patterns, and data repair schemes. It provides real-time knowledge query and reasoning support for other layers and is the foundation for the system to achieve intelligence.

[0018] In the above embodiments, standardized dedicated line routing data is obtained by standardizing the original dedicated line routing data. Preprocessing of the standardized dedicated line routing data yields preprocessed dedicated line routing text. Routing optimization database, routing repair database, and original device parameters are obtained from a preset domain knowledge base. Target resource object information is obtained by identifying and analyzing the preprocessed dedicated line routing text based on the original device parameters. Fault analysis is performed on the target resource object information based on business intent instructions, the routing optimization database, and the routing repair database. Routing repair results are generated based on the analysis results. This solves the problems of automatically converting unstructured raw information into high-quality network resource data, automatically realizing a complete closed loop of handling after route construction failure from root cause diagnosis, task dispatch, repair execution to result verification, and intelligent reasoning and decision-making based on domain knowledge in complex and uncertain scenarios, driving automatic collaborative operation of multiple heterogeneous systems. It eliminates the high dependence on manual input, improves the intelligence level and accuracy of data source processing, replaces the current inefficient manual offline supervision mode, significantly shortens the problem-solving cycle, and improves the adaptability, reliability, and overall intelligence level of the route construction process.

[0019] Optionally, as an embodiment of the present invention, the process of preprocessing each of the standardized dedicated line traffic data in the intelligent core layer to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data includes: Each of the original dedicated line traffic data is cleaned to obtain cleaned dedicated line traffic data corresponding to each of the original dedicated line traffic data. Each of the cleaned dedicated line route data is encoded to obtain the original dedicated line route text corresponding to each of the original dedicated line route data; Each of the original dedicated line route texts is segmented into words to obtain segmented dedicated line route texts corresponding to each of the original dedicated line route data. Each segmented dedicated line route text is tagged with part-of-speech tags to obtain preprocessed dedicated line route texts corresponding to each original dedicated line route data.

[0020] Specifically, the engine first performs format cleaning and unified encoding on the received raw data (i.e., standardized dedicated line routing data). Based on the input text data (such as construction ledgers, completion reports, etc.), it performs basic natural language processing operations such as word segmentation and part-of-speech tagging. After preprocessing, all information is integrated into a unified corpus containing text fragments, key identifiers, and location information for use in subsequent recognition steps.

[0021] In the above embodiments, the standardized dedicated line traffic data is preprocessed to obtain preprocessed dedicated line traffic text, which eliminates the high dependence on manual input and improves the intelligence level and accuracy of data source processing.

[0022] Optionally, as an embodiment of the present invention, in the intelligent core layer, the process of identifying and analyzing all the preprocessed dedicated line route texts based on all the original device parameters to obtain target resource object information includes: Based on all the original equipment parameters, all the preprocessed dedicated line route texts are identified to obtain the original resource object information; The original resource object information is mapped to obtain the mapped resource object information; The mapped resource object information is associated with the preset resource library ID to obtain the associated resource object information. The associated resource object information is connected according to the preset connector words to obtain the connected resource object information; The linked resource object information is verified based on the resource object information in the preset resource library, and the verified information is marked to obtain the marked resource object information. The marked resource object information is supplemented according to the preset topology rules to obtain the target resource object information.

[0023] It should be understood that the core function of the intelligent core layer is to process unstructured information. It calls upon the domain knowledge base (i.e., the preset domain knowledge base), uses technologies such as natural language processing, automatically performs entity recognition, extracts devices, ports, etc. from text, performs semantic association and standardization (such as mapping "0-4-5" to "GigabitEthernet0 / 4 / 5"), conflict detection and intelligent completion, and finally outputs high-quality, structured resource objects (i.e. target resource object information) that can be directly used for routing calculation.

[0024] Specifically, the engine invokes a domain knowledge base (i.e., a pre-defined domain knowledge base), utilizing pre-stored professional knowledge such as network device model libraries, port naming convention libraries, and data center and station name tables to intelligently identify the pre-processed corpus in conjunction with contextual semantics (i.e., the pre-processed leased line routing text). For example, for a description in the construction text "Jiulongpo Zouma Digital Fixed-line Mobile-ATN980C-GigabitEthernet0-4-5", the engine identifies "ATN980C" as a "device", confirms its complete model as "ATN980C-SM-A1" through the knowledge base, identifies "GigabitEthernet0-4-5" as a "port", captures its attribution relationship with the device "ATN980C", identifies "Jiulongpo Zouma Digital Fixed-line Mobile" as a "data center / station", and associates it with the standard geographical information in the knowledge base.

[0025] Understandably, based on the identification of discrete entities, non-standard, colloquial expressions (i.e., mapped resource object information) are mapped into standard formats and encodings that can be recognized by various IT systems (such as resource management systems, network management systems, etc.). For example, port notations from different manufacturers and with different conventions, such as "0-4-5", "gei-0 / 4 / 5", and "GigabitEthernet0 / 4 / 5", are uniformly mapped to the standard port name "GigabitEthernet0 / 4 / 5" for the device in the resource database, and associated with its unique resource instance ID. Next, relationships between entities are constructed based on conjunctions in the text (such as "connect", "link", "peer end", etc.). For example, from "ATN980C's 0-4-5 port connects to W1186's 7th port", a connection relationship (i.e., connected resource object information) is constructed between "device A's port P1" and "device B's port P2".

[0026] Specifically, the engine verifies the standardized entities and their relationships (i.e., the connected resource object information) produced in the above steps against the existing data in the resource library (i.e., the preset resource library). If the extracted port type, speed, or other information is found to be inconsistent with the records in the resource library, it is marked as a "data conflict" and an anomaly is recorded (i.e., the marked resource object information). For some missing information that can be deduced from the context and network topology rules, the engine automatically completes it. For example, if the port connection relationship between the two devices is clear, but the optical path resource is missing, the engine can automatically generate an optical path object based on the relay connection rules in the knowledge base. After the above steps, the engine finally outputs one or more structured resource objects (i.e., target resource object information) and strictly follows the data format instance of JSON Schema. Taking the construction text "Jiulongpo Zouma Data Fixed-line Mobile-ATN980C-GigabitEthernet0-4-5 connected to W1186-Jiulongpo Zouma Data Center-GE2 / 1" in the embodiment as an example, the structured resource object output after engine processing includes the following two core objects: Object 1: Port resource object, belonging to the device "Jiulongpo Zouma Data Fixed-line Mobile-ATN980C", and Object 2: Connection relationship object, describing the connection between the two ports.

[0027] It should be understood that this invention is not a general natural language processing approach, but rather a deep integration of a knowledge base in the communications domain (including device model aliases, port naming rules, topology connection logic, etc.). Through a collaborative model of domain entity recognition, contextual semantic association, and conflict detection, it automatically and accurately converts and completes free text such as "ATN980C's 0-4-5 ports connect to W1186's 7th port in the computer room" into a complete structured resource object that meets the requirements of a resource management system. The core of this method lies in the semantic understanding and structured conversion rules driven by domain knowledge, solving the bottleneck of existing technologies that rely entirely on manual information translation and input. It represents an innovative application within a specific technical field.

[0028] In the above embodiments, target resource object information is obtained by identifying and analyzing all preprocessed dedicated line route texts based on all original equipment parameters. This solves the bottleneck of relying entirely on manual information translation and input in the prior art, eliminates the high dependence on manual input, and improves the intelligence level and accuracy of data source processing.

[0029] Optionally, as an embodiment of the present invention, the process of performing fault analysis on the target resource object information according to the business intent instruction, the route optimization database and the route repair database in the intelligent core layer, and generating route repair results based on the analysis results includes: Based on the business intent instruction and the routing optimization database, the target resource object information is diagnosed and analyzed to obtain a structured diagnostic report; The structured diagnostic report is analyzed and repaired based on the route repair database to obtain the route repair results.

[0030] It should be understood that the core task of receiving standardized structured resource objects (i.e., target resource object information) is to combine these static resource data with dynamic business intentions and system states, transform them into an executable task flow, and drive the underlying existing system to complete route construction.

[0031] In the above embodiments, fault analysis is performed on the target resource object information based on the business intent instruction, the route optimization database, and the route repair database, and route repair results are generated based on the analysis results. This eliminates the high dependence on manual input and improves the intelligence level and accuracy of data source processing.

[0032] Optionally, as an embodiment of the present invention, the routing optimization database includes a prompt word library, a standardized identifier library, a business process knowledge graph, a strategy template library, an atomic task template library, an error code mapping table, and a fault diagnosis knowledge graph; The process of performing diagnostic analysis on the target resource object information based on the business intent instruction and the routing optimization database to obtain a structured diagnostic report includes: Based on the business intent instruction, the preset large model and the prompt word library, semantic matching is performed on the target resource object information to obtain the resource object language and resource object type; The resource object identifier is obtained by mapping the resource object language and the resource object type according to the standardized identifier library. The resource object identifier is decomposed based on the business process knowledge graph to obtain multiple resource object tasks; Based on the business process knowledge graph, multiple task constraints are extracted from all the resource object tasks. Based on the strategy template library, strategy matching is performed on all the resource object tasks and all the task constraints to obtain multiple task execution strategies; Based on the atomic task template library, all the task execution strategies are matched to obtain an atomic task queue; The atomic task queue is routed to obtain multiple circuit execution states; If the circuit execution state is a failure state, the circuit execution state is encapsulated to obtain a failure event object, thereby obtaining multiple failure event objects; Extract the original error code corresponding to each failure event object from each failure event object; The original error codes are mapped according to the error code mapping table to obtain the circuit problem type corresponding to each failure event object. Each failure event object is enhanced using a preset query system to obtain context information corresponding to each failure event object. Based on the fault diagnosis knowledge graph, fault diagnosis is performed on each of the context information to obtain fault information corresponding to each of the failure event objects; A structured diagnostic report is generated based on all the fault information and all the circuit problem types.

[0033] Specifically, the intelligent scheduling center first receives two types of inputs: upstream inputs are from a standardized circuit data list (i.e., target resource object information), which includes a list of circuits to be processed and their attributes; and dynamic inputs are from business intent instructions, such as "execute route import for circuit A", "initiate audit for circuit B", "please help me connect these circuits, with business identifiers A, B, C, etc.", and the underlying system load status is sensed in real time, such as the queue length of the existing route calculation system and API response latency. Based on this, the intelligent scheduling center performs intent parsing and strategy generation.

[0034] Understandably, the scheduling center first performs semantic parsing on the original text input by the user (i.e., the target resource object information) to identify the core intent type of the instruction. This process is based on the natural language understanding capabilities of the large model (i.e., the pre-set large model) and is matched with the prompt word project (i.e., the prompt word library) in the domain knowledge base.

[0035] As should be understood, as shown in Table 1, after clarifying the intent type, the scheduling center performs refined identification and parameterization of the business entities involved in the instruction (i.e., resource object language and resource object type). This step also calls the domain knowledge base (i.e., the standardized identifier library) to map the references in natural language to standardized identifiers (i.e., resource object identifiers) that the system can recognize. Table 1 shows the standardized identifier library.

[0036] Table 1 Specifically, after identifying intent and entities, the scheduling center decomposes the macroscopic user instructions (i.e., resource object identifiers) into multiple quantifiable sub-goals (i.e., resource object tasks) and extracts various constraints affecting task execution (i.e., task constraints). This step relies on the business process knowledge graph in the domain knowledge base. For example, for the input "Please help me assemble circuits A, B, and C", the knowledge graph is queried to find that "assembly" includes three sub-processes. A standardized task template is generated for each circuit. Then, implicit constraints are extracted. There are no explicit priority requirements, and the default sequential scheduling principle is adopted. For the input "Please prioritize assembling these three circuits A, B, and C for government and enterprise customer Zhang San, and then process the others", the extracted constraints are first decomposed into two sub-goals: sub-goal 1 is to immediately assemble circuits A, B, and C, and sub-goal 2 is to assemble the other circuits that the system is currently executing later.

[0037] As should be understood, as shown in Table 2, after clarifying the sub-goals and constraints, the scheduling center enters the core strategy generation stage. Dynamic search and combination optimization are performed on the strategy templates defined by the domain knowledge base (i.e., the strategy template library). The scheduling center inputs the sub-goals and constraints produced by the third layer as query conditions into the strategy space. Then, it retrieves all matching strategy templates in the knowledge base, prioritizes the retrieved strategy templates, instantiates the selected strategy templates with the real-time perceived system status (current API response latency, queue length, resource lock status, etc.) as parameters, and finally outputs the dynamically generated, executable strategy set (i.e., task execution strategy). Table 2 shows the strategy template library.

[0038] Table 2 Specifically, as shown in Table 3, after the strategy is generated, the scheduling center enters the atomic task chain orchestration and queue generation stage. The core task of this step is to transform the output dynamic execution strategy set (i.e., task execution strategy) into a task queue (i.e., atomic task queue) that can be atomically executed by the Agent cluster. First, the scheduling center queries the pre-stored atomic task template library in the domain knowledge base. The atomic task template library matches a corresponding task template for each operation to be executed. For the three circuits (A, B, and C) in this example, the scheduling center matches a standard splicing task chain template for each circuit. This template contains three sequentially executed atomic tasks: route import (ROUTE_IMPORT), resource audit (AUDIT_QUERY), and start splicing (SPLICE_START). Next, the scheduling center injects the execution strategy set generated in the fourth layer (including priority rules, batching plans, concurrency control, retry rules, etc.) into each atomic task, assigning specific execution attributes and constraints to each task instance. After instantiation and constraint injection, the scheduling center organizes all atomic tasks into an ordered atomic task queue according to the principle of priority first and then sorting the same priority according to the dependency relationship. Table 3 is the atomic task template library.

[0039] Table 3 It should be understood that after the atomic task queue is generated, it will be submitted to the atomic execution agent cluster. Each agent in the cluster determines when to dispatch the task to the corresponding execution agent based on the scheduled execution time (scheduledTime) and the current concurrency.

[0040] As should be understood, as shown in Table 4, failure signals are captured in three ways. The framework first encapsulates the original failure information into a unified failure event object for easy subsequent processing. Table 4 is the library for capturing failure signals.

[0041] Table 4 Specifically, upon receiving a standardized failure event, the following four sub-steps are performed for root cause inference: a. Error code classification and preliminary location As shown in Table 5, the diagnostic agent first queries the error code mapping table in the domain knowledge base to map the original error code to the problem type. Table 5 is the error code mapping table.

[0042] Table 5 b. Enhanced contextual information The diagnostic agent extracts context snapshots from failure events (i.e., failure event objects) and actively inquires about the current status of the relevant systems to obtain more diagnostic information. In the example, the diagnostic agent first inquires the resource repository (i.e., the preset inquiry system) to confirm whether the device "Wantong 150X" does not exist, or whether it exists but is in an abnormal state. Then, it inquires the work order system (i.e., the preset inquiry system) to confirm whether there are pending or completed work orders related to the device. Finally, it inquires the historical records (i.e., the preset inquiry system) to retrieve historical failure records of the circuit or the device and check for similar patterns.

[0043] c. Domain knowledge graph reasoning As shown in Table 6, the diagnostic agent invokes the fault diagnosis knowledge graph in the domain knowledge base for deep reasoning. Various fault modes, root causes and responsible party mapping relationships of various faults are pre-stored in the knowledge graph. In Table 6, based on the error code DEVICE_NOT_FOUND and the device name "Wantong 150X", the "user-side device missing" mode is matched, the root cause output is "The customer-side device 'Wantong 150X' is not registered in the resource repository", the responsible party output is "local branch", and the diagnostic agent finally outputs a structured diagnostic report (i.e., structured diagnostic report). Table 6 is the fault diagnosis knowledge graph.

[0044] Table 6 In the above embodiment, the target resource object information is diagnosed and analyzed according to the business intent instruction and the routing optimization database to obtain a structured diagnostic report, which improves the intelligence and accuracy of data source processing, replaces the current inefficient manual offline supervision mode, significantly shortens the problem-solving cycle, and improves the adaptability, reliability and overall intelligence of the route construction process.

[0045] Optionally, as an embodiment of the present invention, the route repair database includes a repair policy base and a verification rule base; The process of performing repair analysis on the structured diagnostic report according to the route repair database to obtain a route repair result includes: Acquiring an initial repair task chain from the repair policy base according to the structured diagnostic report; Updating parameters of the initial repair task chain according to the structured diagnostic report to obtain an updated repair task chain, and using the updated repair task chain as the route repair result.

[0046] It should be understood that, after receiving the diagnostic report (i.e., structured diagnostic report), dynamic orchestration based on the repair policy base will be performed to generate a specific repair task chain (i.e., initial repair task chain).

[0047] Specifically, the domain knowledge base pre-stores a repair strategy library, which contains standardized repair templates for different root cause types. The scheduling agent matches the corresponding template from the repair strategy library based on rootCause.type and targetObject in the diagnostic report, and dynamically fills the template (i.e., the initial repair task chain) with the parameters in the diagnostic report (i.e., the structured diagnostic report). In this example, rootCauseType is "RESOURCE_MISSING" and targetObject is "device", matching the template REPAIR_DEVICE_MISSING to generate a repair task chain (i.e., the updated repair task chain).

[0048] In the above embodiments, the structured diagnostic report is repaired and analyzed based on the route repair database to obtain the route repair results. This replaces the current inefficient manual offline supervision mode, significantly shortens the problem-solving cycle, and improves the adaptability, reliability, and overall intelligence level of the route construction process.

[0049] Optionally, as an embodiment of the present invention, the intelligent core layer is further configured to: S71: Create an execution work order based on the route repair results, and execute all tasks in the execution work order; S72: Obtain the work order status from the preset scheduling system; S73: If the work order status is the first preset status, then generate the first preset work order completion result, use the first preset work order completion result as the work order completion result, and execute S76. S74: If the work order status is the second preset status or the work order status is the third preset status, then generate the second preset work order completion result, use the second preset work order completion result as the work order completion result, and execute S76. S75: If the preset stop time is reached, a third preset work order completion result is generated, and the third preset work order completion result is used as the work order completion result, and S76 is executed. S76: Obtain the device status from the preset resource center device, and generate a verification report based on the work order completion result and the device status; S77: If the verification report indicates that the verification passed, then the target resource object information is re-analyzed based on the business intent instruction, the route optimization database, and the route repair database; if the verification report indicates that the verification failed, then an exception message is generated.

[0050] It should be understood that the first preset work order completion result can be work order completion successful, the second preset work order completion result can be work order completion failure, and the third preset work order completion result can be work order completion timeout.

[0051] Specifically, this is accomplished by the executing intelligent agent, which is responsible for sequentially calling the API interfaces of various external systems to perform specific operations according to the definition of the repair task chain. As shown in the example, it involves the following steps: A. Create a follow-up work order Call the API to create the following information, and execute the agent to extract the ticketId. B. Monitor work order status The intelligent agent initiates a polling task, querying the scheduling system for the work order status every second.

[0052] Polling continues until one of the following conditions is met: Success condition: The work order status changes to "CLOSED", indicating that the branch office has completed the data entry; Timeout condition: If the task is not completed within 24 hours, the onFailure action task will be triggered. Failure condition: The work order status changes to "REJECTED" or "CANCELLED", triggering the onFailure action task.

[0053] Suppose that after 1 hour, the polling returns the following result: the agent determination was successful.

[0054] C. Verify whether the resources have been added. The agent calls the device query API of the resource center (i.e., the preset resource center device) to verify whether the device (i.e. the device status) already exists. The agent then executes the application verification rule response.data.total>0 and response.data.items[0].status == 'ACTIVE', and the verification is deemed successful.

[0055] As should be understood, as shown in Table 7, the execution agent completes the task, and the verification agent is responsible for the final verification of the overall execution result of the repair task chain and triggering subsequent actions based on the result. The verification agent outputs a structured verification report (i.e., the verification report). After the verification is passed, the verification agent automatically triggers the subsequent process according to the onSuccess action task defined in the repair task chain and executes the re-trigger interface. After receiving the request, the corresponding module re-executes the fault analysis of the circuit: at this time the device already exists, and the splicing is expected to be successfully completed. Table 7 is the verification content table.

[0056] Table 7 In the above embodiments, the verification and analysis of the route repair results significantly shortened the problem-solving cycle and improved the adaptability, reliability, and overall intelligence level of the route construction process.

[0057] Alternatively, as another embodiment of the present invention, such as Figure 2 As shown, this invention includes a module (101) for data access, an intelligent parsing and data governance engine (102) for implementing key point 1, a routing construction and optimization core (103) for implementing key point 2, a closed-loop self-healing execution framework (104) for implementing key point 3, a console (105) providing an interactive interface, and a domain knowledge base (106) as core support. This invention protects the specific system architecture and the collaborative relationships between its modules, specifically protecting the specific technical system structure composed of modules 102, 103, 104, and 106, capable of realizing the process from unstructured data input to route construction completion, and achieving intelligent diagnosis and closed-loop self-healing in case of anomalies. The protection points of this invention focus on how to utilize large models, multiple agents, and domain knowledge bases in the specific field of government and enterprise dedicated line routing to solve the three specific technical problems of intelligent data governance, intelligent process scheduling, and cross-system intelligent closed-loop, and the collaborative relationships between them. These points constitute the core technological innovation that distinguishes it from existing general automation solutions or single-point tools.

[0058] Optionally, as another embodiment of the present invention, a framework is designed in which diagnostic, scheduling, execution, and verification agents work collaboratively. Upon receiving a failure code, the diagnostic agent does not simply match the error code, but instead performs root cause reasoning by combining the fault context, system state snapshots, and domain knowledge base, such as inferring that "the user-end device has not been recorded" and the responsible party is "Branch XX". The scheduling agent then generates a specific repair task chain across systems, such as "pushing a DingTalk supervision message to Branch XX's resource supplementation work order". The execution agent drives the execution of each system interface, and finally, the verification agent confirms the repair and triggers a retry, forming a closed loop.

[0059] Optionally, as another embodiment of the present invention, the present invention introduces a large model-driven intelligent scheduling hub, which can achieve the following: (i) understanding high-level business intentions or performance goals such as "Please help me assemble these circuits, the business identifier is XXXX"; (ii) dynamically generating and optimizing the execution strategy, priority, and timing arrangement of atomic tasks (such as route import, audit, assembly, etc.). Subsequently, through an orchestratable atomic execution agent cluster (such as route import agent, audit trigger agent, etc.), the task is executed by strictly following this strategy and calling the standard API interface of the existing system.

[0060] Alternatively, as another embodiment of the present invention, the key point of the present invention lies in creatively combining a large language model, a multi-agent collaborative framework, and the standard capabilities of existing telecommunications operation and maintenance systems to construct an intelligent device covering the entire link of data governance, intelligent scheduling, and closed-loop self-healing.

[0061] Optionally, as another embodiment of the present invention, the present invention primarily addresses the intelligent governance and closed-loop management of network resource data in the context of building dedicated enterprise and government lines for telecom operators. The proposed method deeply integrates specific knowledge from this vertical field, such as circuit models, audit rules, and cross-domain processes. For network resource data governance problems in other fields, such as IP network configuration management and cloud resource orchestration, although the general goals of automation and intelligence are similar, the specific technical solutions and system structures described in this invention cannot be directly applied or require substantial modification due to fundamental differences in domain knowledge, system architecture, and business processes.

[0062] Optionally, as another embodiment of the present invention, the objectives are: first, to solve the problem of automatically converting unstructured raw information into high-quality network resource data, thereby eliminating the heavy reliance on manual input and improving the intelligence and accuracy of data source processing; second, to solve the problem of automatically implementing a complete closed-loop handling process after route construction failure, from root cause diagnosis, task assignment, repair execution to result verification, replacing the current inefficient manual offline supervision mode and significantly shortening the problem-solving cycle; and third, to solve the problem of the system performing intelligent reasoning and decision-making based on domain knowledge in complex and uncertain scenarios, and driving the automatic collaborative operation of multiple heterogeneous systems, thereby improving the adaptability, reliability, and overall intelligence level of the route construction process.

[0063] Alternatively, as another embodiment of the present invention, the core of the present invention lies in using large model and multi-agent technology to intelligently orchestrate and automatically schedule existing dispersed resources and processes, thereby achieving a closed loop from data governance and route construction to fault self-healing.

[0064] Alternatively, as another embodiment of the present invention, such as Figure 3 As shown, the present invention includes: S201: Multi-source data access and event triggering. The process is triggered upon capturing events such as the submission of a new completion report or the start of a scheduled task.

[0065] S202: Intelligent Analysis and Governance of Raw Data. Automatically analyzes, standardizes, and performs conflict checks on unstructured construction text and other data from events, outputting structured circuit resource data.

[0066] S203: Intelligent Scheduling and Route Construction Execution. The core steps specifically include: The intelligent scheduling hub dynamically formulates task execution strategies based on business objectives and S202 outputs, combined with a knowledge base. The orchestration engine decomposes the strategy into atomic task chains such as "route import, circuit audit, and startup splicing." Next, the atomic execution agent cluster is driven, sequentially calling the standard APIs of the existing system to complete the entire route construction process. All core logic, such as route calculation and rule verification, is handled by the underlying existing system. Finally, the final execution status of all circuits (success / failure and reason) is summarized, and the process proceeds to the judgment phase.

[0067] S204: Route construction result determination. The system determines the branch based on the summary results of S203.

[0068] S205: Successful Result Verification and Process End. If all steps are successful, the process ends, and the successful result is notified to the relevant personnel.

[0069] S206: Failure Result Determination and Root Cause Intelligent Diagnosis. If a failure exists, diagnosis is initiated. The diagnostic agent analyzes the failure data, combines it with the knowledge base, and outputs an accurate root cause diagnosis report, such as "Equipment A was not entered; the responsible party is branch company B".

[0070] S207: Generate and execute a self-healing task chain. The scheduling agent generates a specific task chain based on the diagnostic report, such as "pushing a DingTalk supervision message to team B," which is then executed by the execution agent through the external system interface.

[0071] S208: Self-healing effect verification and closed loop. After the verification agent confirms that the repair is completed, the system automatically jumps back to S203 and re-initiates route construction, forming an intelligent closed loop.

[0072] Optionally, as another embodiment of the present invention, the present invention includes: Scenario: 300 existing circuit splices need to be processed daily. The existing system executes them in a fixed order, taking about 8 hours.

[0073] Applying this invention: In S203, the scheduling center analyzes the circuit characteristics and dynamically formulates strategies based on the intention of "Please help me assemble these circuits, the service identifier is XXXX". Prioritizes processing 200 circuits with simple topologies to quickly obtain a success rate; for 100 circuits with a high historical failure rate, a strategy of batch execution and retrying at intervals after failure is adopted.

[0074] Results: Based on the large model capability, the agent intelligent orchestration avoids resource contention and invalid retries, reducing the total time to 4.5 hours and increasing the success rate from about 70% to 96%.

[0075] Optionally, as another embodiment of the present invention, the present invention includes: Scenario: Circuit “831JPW21762350” failed in the splicing task of S203, returning the error “User device not recorded”.

[0076] S206 Diagnosis: The diagnostic agent determined that the missing device was "Anhui Tong 150X". Based on the rule in the knowledge base that "customer-side devices are entered by the local branch company", the responsible party was identified as "XX Branch Company".

[0077] S207 Execution: The scheduling agent generates a task chain, and the execution agent automatically calls the scheduling system API to create a work order with the title "Please supplement the user-side equipment 'Anhui Tong 150X' for circuit 831JPW21762350" and notify the corresponding position in XX branch.

[0078] S208 Verification and Closed Loop: Once the system detects the work order completion, the verification agent queries the resource library to confirm that the device has been entered, and then automatically re-triggers the S203 process for that circuit, ultimately achieving successful assembly. The entire process requires no manual searching, communication, or manual retries.

[0079] Figure 4 This is a flowchart illustrating a dedicated line traffic data management method according to an embodiment of the present invention.

[0080] Alternatively, as another embodiment of the present invention, such as Figure 4 As shown, a dedicated line traffic data management method includes the following steps: S1: Collect multiple raw dedicated line routing data, and standardize each raw dedicated line routing data to obtain standardized dedicated line routing data corresponding to each raw dedicated line routing data; S2: Preprocess each of the standardized dedicated line traffic data to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data; S3: Obtain the route optimization database, route repair database, and multiple raw device parameters from the preset domain knowledge base; S4: Based on all the original equipment parameters, identify and analyze all the preprocessed dedicated line route texts to obtain target resource object information; S5: Import business intent instructions, perform fault analysis on the target resource object information based on the business intent instructions, the route optimization database, and the route repair database, and generate route repair results based on the analysis results.

[0081] Optionally, another embodiment of the present invention provides a dedicated line traffic data management system, 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, it implements the dedicated line traffic data management method as described above. This system can be a computer or similar system.

[0082] Optionally, another embodiment of the present invention provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the leased line routing data management method as described above.

[0083] It should be noted that, in this document, relational terms such as "first" and "second" are used only to distinguish one entity or operation from another, and do not necessarily require or imply any such actual relationship or order between these entities or operations. Furthermore, the terms "comprising," "including," or any other variations thereof are intended to cover non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements includes not only those elements but also other elements not expressly listed, or elements inherent to such process, method, article, or apparatus.

[0084] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the specific working process of the above-described apparatus and unit can be referred to the corresponding process in the foregoing method embodiments, and will not be repeated here.

[0085] In the several embodiments provided in this application, it should be understood that the disclosed apparatus and methods can be implemented in other ways. For example, the apparatus embodiments described above are merely illustrative. For instance, the division of units is only a logical functional division, and in actual implementation, there may be other division methods. For example, multiple units or components may be combined or integrated into another system, or some features may be ignored or not executed.

[0086] The units described as separate components may or may not be physically separate. The components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the units can be selected to achieve the purpose of the embodiments of the present invention, depending on actual needs.

[0087] Furthermore, the functional units in the various embodiments of the present invention can be integrated into one processing unit, or each unit can exist physically separately, or two or more units can be integrated into one unit. The integrated unit can be implemented in hardware or as a software functional unit.

[0088] If the integrated unit is implemented as a software functional unit and sold or used as an independent product, it can be stored in a computer-readable storage medium. This is understood to mean that the technical solution of the present invention, in essence, or the part that contributes to the prior art, or all or part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0089] The above are merely specific embodiments of the present invention, but the scope of protection of the present invention is not limited thereto. Any person skilled in the art can easily conceive of various equivalent modifications or substitutions within the technical scope disclosed in the present invention, and these modifications or substitutions should all be covered within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A dedicated line route data management system, characterized in that, include: Data access layer, intelligent core layer, support layer, and application interaction layer. The data access layer is used to collect multiple raw dedicated line traffic data, and to standardize each raw dedicated line traffic data to obtain standardized dedicated line traffic data corresponding to each raw dedicated line traffic data. The intelligent core layer is used to preprocess each of the standardized dedicated line traffic data to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data. The support layer is used to obtain a route optimization database, a route repair database, and multiple raw device parameters from a preset domain knowledge base; The intelligent core layer is also used to identify and analyze all the preprocessed dedicated line route texts based on all the original equipment parameters to obtain target resource object information; The application interaction layer is used to import business intent commands; The intelligent core layer is also used to perform fault analysis on the target resource object information according to the business intent instruction, the route optimization database and the route repair database, and generate route repair results based on the analysis results.

2. The dedicated line route data management system according to claim 1, characterized in that, In the intelligent core layer, the process of preprocessing each of the standardized dedicated line traffic data to obtain the preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data includes: Each of the original dedicated line traffic data is cleaned to obtain cleaned dedicated line traffic data corresponding to each of the original dedicated line traffic data. Each of the cleaned dedicated line route data is encoded to obtain the original dedicated line route text corresponding to each of the original dedicated line route data; Each of the original dedicated line route texts is segmented into words to obtain segmented dedicated line route texts corresponding to each of the original dedicated line route data. Each segmented dedicated line route text is tagged with part-of-speech tags to obtain preprocessed dedicated line route texts corresponding to each original dedicated line route data.

3. The dedicated line route data management system according to claim 1, characterized in that, In the intelligent core layer, the process of identifying and analyzing all preprocessed dedicated line route texts based on all the original device parameters to obtain target resource object information includes: Based on all the original equipment parameters, all the preprocessed dedicated line route texts are identified to obtain the original resource object information; The original resource object information is mapped to obtain the mapped resource object information; The mapped resource object information is associated with the preset resource library ID to obtain the associated resource object information. The associated resource object information is connected according to the preset connector words to obtain the connected resource object information; The linked resource object information is verified based on the resource object information in the preset resource library, and the verified information is marked to obtain the marked resource object information. The marked resource object information is supplemented according to the preset topology rules to obtain the target resource object information.

4. The dedicated line route data management system according to claim 1, characterized in that, In the intelligent core layer, the process of performing fault analysis on the target resource object information according to the business intent instruction, the route optimization database, and the route repair database, and generating route repair results based on the analysis results includes: Based on the business intent instruction and the routing optimization database, the target resource object information is diagnosed and analyzed to obtain a structured diagnostic report; The structured diagnostic report is analyzed and repaired based on the route repair database to obtain the route repair results.

5. The dedicated line route data management system according to claim 4, characterized in that, The routing optimization database includes a prompt word library, a standardized identifier library, a business process knowledge graph, a strategy template library, an atomic task template library, an error code mapping table, and a fault diagnosis knowledge graph. The process of performing diagnostic analysis on the target resource object information based on the business intent instruction and the routing optimization database to obtain a structured diagnostic report includes: Based on the business intent instruction, the preset large model and the prompt word library, semantic matching is performed on the target resource object information to obtain the resource object language and resource object type; The resource object identifier is obtained by mapping the resource object language and the resource object type according to the standardized identifier library. The resource object identifier is decomposed based on the business process knowledge graph to obtain multiple resource object tasks; Based on the business process knowledge graph, multiple task constraints are extracted from all the resource object tasks. Based on the strategy template library, strategy matching is performed on all the resource object tasks and all the task constraints to obtain multiple task execution strategies; Based on the atomic task template library, all the task execution strategies are matched to obtain an atomic task queue; The atomic task queue is routed to obtain multiple circuit execution states; If the circuit execution state is a failure state, the circuit execution state is encapsulated to obtain a failure event object, thereby obtaining multiple failure event objects; Extract the original error code corresponding to each failure event object from each failure event object; The original error codes are mapped according to the error code mapping table to obtain the circuit problem type corresponding to each failure event object. Each failure event object is enhanced using a preset query system to obtain context information corresponding to each failure event object. Based on the fault diagnosis knowledge graph, fault diagnosis is performed on each of the context information to obtain fault information corresponding to each of the failure event objects; A structured diagnostic report is generated based on all the fault information and all the circuit problem types.

6. The dedicated line route data management system according to claim 4, characterized in that, The route repair database includes a repair policy library and a verification rule library; The process of performing repair analysis on the structured diagnostic report based on the route repair database to obtain the route repair results includes: The initial repair task chain is obtained from the repair strategy library based on the structured diagnostic report; The parameters of the initial repair task chain are updated based on the structured diagnostic report to obtain the updated repair task chain, and the updated repair task chain is used as the routing repair result.

7. The dedicated line route data management system according to claim 1, characterized in that, The intelligent core layer is also used for: S71: Create an execution work order based on the route repair results, and execute all tasks in the execution work order; S72: Obtain the work order status from the preset scheduling system; S73: If the work order status is the first preset status, then generate the first preset work order completion result, use the first preset work order completion result as the work order completion result, and execute S76. S74: If the work order status is the second preset status or the work order status is the third preset status, then generate the second preset work order completion result, use the second preset work order completion result as the work order completion result, and execute S76. S75: If the preset stop time is reached, a third preset work order completion result is generated, and the third preset work order completion result is used as the work order completion result, and S76 is executed. S76: Obtain the device status from the preset resource center device, and generate a verification report based on the work order completion result and the device status; S77: If the verification report indicates that the verification passed, then the target resource object information is re-analyzed based on the business intent instruction, the route optimization database, and the route repair database; if the verification report indicates that the verification failed, then an exception message is generated.

8. A dedicated line route data management method, characterized in that, The steps include the following: Multiple raw dedicated line routing data are collected, and each raw dedicated line routing data is standardized to obtain standardized dedicated line routing data corresponding to each raw dedicated line routing data. Each of the standardized dedicated line traffic data is preprocessed to obtain preprocessed dedicated line traffic text corresponding to each of the original dedicated line traffic data. Obtain route optimization database, route repair database, and multiple raw device parameters from the preset domain knowledge base; Based on all the original equipment parameters, all the preprocessed dedicated line route texts are identified and analyzed to obtain target resource object information; Import business intent instructions, perform fault analysis on the target resource object information based on the business intent instructions, the route optimization database, and the route repair database, and generate route repair results based on the analysis results.

9. A dedicated line data management system, 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, it implements the leased line traffic data management method as described in claim 8.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by the processor, the leased line traffic data management method as described in claim 8 is implemented.

Images