Method and device for implementing commission-dimension intelligent agent based on large model

Abstract

The invention relates to the technical field of communication network operation and maintenance, and particularly provides a method and device for realizing a commission maintenance agent based on a large model, and the method comprises the following steps: S1, generating a fault processing logic model; s2, pushing the multi-dimensional data; s3, intelligent path planning; s4, business influence query; s5, root cause analysis and material recommendation; s6, upper navigation and fault point positioning; s7, repairing scheme recommendation and performance monitoring; and S8, recommending the work order. Compared with the prior art, automation, intelligence and visualization of the whole fault processing process can be realized through cooperation of a large model and digital employees.

Description

Technical Field

[0001] This invention relates to the field of communication network operation and maintenance technology, specifically providing a method and apparatus for implementing a maintenance agent based on a large model. Background Technology

[0002] With the continuous growth of communication network scale, the workload of outsourced maintenance has surged, and the traditional manual processing mode faces the following challenges:

[0003] (1) Low processing efficiency: Fault location, resource allocation, work order execution and other links rely on human experience, resulting in slow response speed and easy errors.

[0004] (2) Difficulty in reusing experience: The skill levels of maintenance personnel vary, making it difficult to systematically accumulate and share historical experience.

[0005] (3) Serious waste of resources: The lack of accurate analysis of the root causes of failures leads to problems such as redundant material preparation and unreasonable path planning.

[0006] (4) Insufficient data collaboration: Data from multiple disciplines (transmission, wireless, environmental, etc.) are fragmented, making it difficult to form a global perspective for decision support.

[0007] While some existing technologies attempt to introduce automation tools, these are mostly single-function modules lacking deep integration with large-scale model technology, thus failing to achieve end-to-end intelligent closed-loop management of outsourced maintenance. Therefore, there is an urgent need for an intelligent outsourced maintenance agent solution based on large-scale models, which can improve the quality and efficiency of outsourced maintenance by replacing or assisting humans with machines. Summary of the Invention

[0008] This invention addresses the shortcomings of the prior art by providing a highly practical method for implementing a large-model-based dimensional agent.

[0009] A further technical objective of this invention is to provide a reasonably designed, safe, and applicable device for implementing a generational intelligent agent based on a large model.

[0010] The technical solution adopted by this invention to solve its technical problem is:

[0011] A method for implementing a dimensional agent based on a large model has the following steps:

[0012] S1. Generate a fault handling logic model;

[0013] S2, push multi-dimensional data;

[0014] S3, Intelligent Path Planning;

[0015] S4. Business Impact Inquiry;

[0016] S5. Root cause analysis and material recommendation;

[0017] S6. Upper navigation and fault location;

[0018] S7. Recommended Repair Solutions and Performance Monitoring;

[0019] S8, Work Order Recommendation.

[0020] Furthermore, in step S1, the fault event type, root cause classification, location data, and urgency level are input, and a graphical processing step model is output. The model is trained using historical fault data to learn the processing logic under different scenarios, generate an interactive graphical interface, and guide maintenance personnel to operate step by step.

[0021] Furthermore, in step S2, resource data push and performance data push are included. The big model, combined with digital employees, collects and generates structured data in real time, and pushes it to maintenance personnel through a visual interface or mobile terminal.

[0022] Furthermore, in step S3, the location of the alarm network element and the real-time location of the maintenance personnel are input into the model, and the optimal on-site path and estimated duration are output.

[0023] The large model calls the map API, combines real-time traffic data with historical route optimization algorithms, dynamically plans routes, and provides navigation guidance.

[0024] Furthermore, in step S4, the alarm event is input, and the affected business scope, number of users, and potential economic losses are output. The large model analyzes the business topology data associated with the alarm, comprehensively assesses the impact, and generates a briefing.

[0025] Furthermore, in step S5, the root cause analysis involves extracting the original alarm location information from the large model, analyzing the cause of the alarm, and generating a processing solution.

[0026] The material preparation involves recommending the required materials based on the fault type and historical repair data, and providing a material application interface.

[0027] The large model uses a knowledge graph to associate fault types with a material library and generates a recommendation list.

[0028] Furthermore, in step S6, OTDR ranging technology is used to directly locate the fault point, provide a navigation path, locate the specific device location through the device SN code, and integrate the GIS system and device database into the large model to provide real-time feedback of location information.

[0029] Furthermore, in step S7, a repair plan is pushed based on the location of the maintenance personnel and the distance to the fault point. The big data model analyzes historical data to predict the repair time and provides real-time feedback on the status through voice broadcast.

[0030] Furthermore, in step S8, the current location of the maintenance personnel and the urgency of the event are input, and a list of tasks to be processed is output. The large model combines a priority algorithm to dynamically recommend work orders.

[0031] A device for implementing a generational intelligent agent based on a large model includes: at least one memory and at least one processor;

[0032] The at least one memory is used to store a machine-readable program;

[0033] The at least one processor is used to call the machine-readable program to execute a method for implementing a large-model-based agent.

[0034] Compared with existing technologies, the present invention provides a method and apparatus for implementing a large-scale model-based agent, which has the following significant advantages:

[0035] This invention, based on the deep integration of large-scale models and digital employee technology, constructs an intelligent solution for the entire communication network maintenance process, significantly improving operation and maintenance efficiency and quality. Through multimodal data fusion (text, image, time-series data) and dynamic knowledge graphs, the system can accurately locate the root cause of faults and generate repair solutions. Combined with path planning, resource push, and intelligent material recommendation functions, it greatly reduces the need for manual intervention.

[0036] Meanwhile, leveraging the natural language interaction and knowledge reasoning capabilities of large-scale models, the system enables frontline personnel to quickly obtain structured analysis reports and decision-making suggestions, improving the accuracy of fault diagnosis. This solution provides core support for operators to achieve cost reduction, efficiency improvement, and digital transformation by promoting the transformation of network operations and maintenance towards "unmanned and self-healing" processes. Attached Figure Description

[0037] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0038] Figure 1 This is a flowchart illustrating a method for implementing a generational intelligent agent based on a large model. Detailed Implementation

[0039] To enable those skilled in the art to better understand the present invention, the present invention will be further described in detail below with reference to specific embodiments. Obviously, the described embodiments are merely some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0040] The following is a preferred embodiment:

[0041] like Figure 1 As shown in this embodiment, a method for implementing a dimensional agent based on a large model has the following steps:

[0042] S1. Generate a fault handling logic model;

[0043] Input: Fault event type, root cause classification, location data, urgency level;

[0044] Output: A graphical model of the processing steps (such as a flowchart or decision tree).

[0045] The large model is trained using historical fault data to learn the processing logic under different scenarios, generating an interactive graphical interface to guide maintenance personnel to operate step by step.

[0046] S2, push multi-dimensional data;

[0047] This includes resource data push and performance data push;

[0048] Resource data push includes:

[0049] Transmission Specialty: Proactively push alarm information including district / county, time, network element name, equipment model, port, slot information, board model, and cell information;

[0050] Wireless Specialist: Push network element name, board model, BBU model and location, SN code, RRU model and location, base station location, and downlink service volume, etc.

[0051] Performance data push includes:

[0052] Transmission equipment: optical power, alarms, temperature;

[0053] Wireless equipment: total load of switching power supply, temperature, humidity, AC voltage, current, VSWR value, number of satellites searched, etc.

[0054] The large model, combined with digital employees, collects and generates structured data in real time, which is then pushed to maintenance personnel through a visual interface or mobile device.

[0055] S3, Intelligent Path Planning;

[0056] Input: Location of alarm network element, real-time location of maintenance personnel;

[0057] Output: Optimal route to the next station and estimated duration.

[0058] The large model calls the map API, combines real-time traffic data with historical route optimization algorithms (such as the A* algorithm), dynamically plans routes, and provides navigation guidance.

[0059] S4. Business Impact Inquiry;

[0060] Input: Alarm event;

[0061] Output: Scope of affected business, number of users, potential economic losses, etc.

[0062] The large model analyzes the business topology data associated with alarms, comprehensively assesses the impact, and generates a briefing.

[0063] S5. Root cause analysis and material recommendation;

[0064] Root cause segmentation: The large model extracts the original alarm location information, analyzes the causes of alarms (such as hardware failure, configuration error, environmental interference, etc.), and generates a solution.

[0065] Material preparation: Based on the fault type and historical repair data, recommend the required materials (such as optical cables, single boards, tools, etc.) and provide a material application interface.

[0066] The large model uses a knowledge graph to associate fault types with a material library and generates a recommendation list.

[0067] S6. Upper navigation and fault location;

[0068] Fiber optic cable interruption event: Combined with OTDR ranging technology, the fault point can be directly located and a navigation path can be provided;

[0069] Hardware failure: Quickly locate the specific device position using the device's serial number (SN).

[0070] The large model integrates GIS systems and equipment databases, providing real-time feedback on location information.

[0071] S7. Recommended Repair Solutions and Performance Monitoring;

[0072] Based on the location of the maintenance personnel and the distance to the fault point, a repair plan (such as replacing the board, adjusting parameters, etc.) is pushed.

[0073] The large model analyzes historical data to predict repair time, dynamically monitors optical power, alarm indicators, and data center temperature, and provides real-time status feedback through voice broadcast.

[0074] S8, Work Order Recommendation;

[0075] Input: Current location of the maintenance personnel, urgency level of the incident;

[0076] Output: List of tasks to be processed.

[0077] Large models are combined with priority algorithms (such as greedy algorithms or reinforcement learning) to dynamically recommend work orders.

[0078] Based on the above method, a device for implementing a large-scale model-based agent in this embodiment includes: at least one memory and at least one processor;

[0079] The at least one memory is used to store a machine-readable program;

[0080] The at least one processor is used to call the machine-readable program to execute a method for implementing a large-model-based agent.

[0081] The above-described specific embodiments are merely specific examples of the present invention. The patent protection scope of the present invention includes, but is not limited to, the above-described specific embodiments. Any technical solution that conforms to the above-described specific embodiments of the present invention and any appropriate changes or substitutions made by those skilled in the art should fall within the patent protection scope of the present invention.

[0082] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for implementing a dimensional intelligent agent based on a large model, characterized in that, It has the following steps: S1. Generate a fault handling logic model; S2, push multi-dimensional data; S3, Intelligent Path Planning; S4. Business Impact Inquiry; S5. Root cause analysis and material recommendation; S6. Upper navigation and fault location; S7. Recommended Repair Solutions and Performance Monitoring; S8, Work Order Recommendation.

2. The method for implementing a large-model-based dimensional intelligent agent according to claim 1, characterized in that, In step S1, the fault event type, root cause classification, location data, and urgency level are input, and a graphical processing step model is output. The model is trained with historical fault data to learn the processing logic under different scenarios and generate an interactive graphical interface to guide maintenance personnel to operate step by step.

3. The method for implementing a dimensional agent based on a large model according to claim 2, characterized in that, Step S2 includes resource data push and performance data push. The big model, combined with digital employees, collects and generates structured data in real time, and pushes it to maintenance personnel through a visual interface or mobile terminal.

4. The method for implementing a dimensional agent based on a large model according to claim 3, characterized in that, In step S3, the location of the alarm network element and the real-time location of the maintenance personnel are input into the model, and the optimal on-site path and estimated duration are output. The large model calls the map API, combines real-time traffic data with historical route optimization algorithms, dynamically plans routes, and provides navigation guidance.

5. The method for implementing a dimensional agent based on a large model according to claim 4, characterized in that, In step S4, the alarm event is input, and the affected business scope, number of users, and potential economic losses are output. The large model analyzes the business topology data associated with the alarm, comprehensively assesses the impact, and generates a briefing.

6. The method for implementing a dimensional agent based on a large model according to claim 5, characterized in that, In step S5, the root cause analysis involves extracting the original alarm location information from the large model, analyzing the cause of the alarm, and generating a processing solution. The material preparation involves recommending the required materials based on the fault type and historical repair data, and providing a material application interface. The large model uses a knowledge graph to associate fault types with a material library and generates a recommendation list.

7. The method for implementing a dimensional agent based on a large model according to claim 6, characterized in that, In step S6, the fault point is directly located by combining OTDR ranging technology, and a navigation path is provided. The specific location of the equipment is located by the equipment SN code. The large model integrates the GIS system and the equipment database to provide real-time feedback of location information.

8. The method for implementing a dimensional agent based on a large model according to claim 7, characterized in that, In step S7, a repair plan is pushed based on the location of the maintenance personnel and the distance to the fault point. The big data model analyzes historical data to predict the repair time and provides real-time feedback on the status through voice broadcast.

9. The method for implementing a dimensional agent based on a large model according to claim 8, characterized in that, In step S8, the current location of the maintenance personnel and the urgency of the event are input, and a list of tasks to be processed is output. The big model dynamically recommends work orders based on a priority algorithm.

10. A device for implementing a generational intelligent agent based on a large model, characterized in that, include: At least one memory and at least one processor; The at least one memory is used to store a machine-readable program; The at least one processor is configured to invoke the machine-readable program to perform the method according to any one of claims 1 to 9.

Images