A fault work order maintenance management method, device, equipment and medium

Abstract

This invention discloses a maintenance management method, device, equipment, and medium for fault work orders, relating to the field of work order operation and maintenance management technology. By determining the maintenance status of faulty server equipment, this invention allows only in-warranty and extended-warranty equipment to create work orders, restricting the creation of fault work orders for out-of-warranty equipment, thus avoiding invalid work order dispatch and unnecessary maintenance costs, and reducing enterprise operation and maintenance costs. A dual-channel fault determination method, which weights and fuses the deviation of equipment operating parameters from their status and the semantic similarity of the fault description text, solves the problem of high misjudgment rate in single-dimensional fault judgment, accurately determining the fault type. Based on the fault type, a suitable maintenance plan is determined, improving maintenance efficiency. Matching fault types with candidate engineer data enables refined allocation of human resources, reducing idle or overloaded human resources. The fault work order and the suitable maintenance plan are sent together to the engineer's bound terminal device, shortening maintenance time and improving customer experience.

Description

Technical Field

[0001] This invention relates to the field of work order operation and maintenance management technology, and in particular to a method, device, equipment and medium for the maintenance management of fault work orders. Background Technology

[0002] The existing method for handling fault work orders involves: collecting IoT data, resource data, and personnel data through a data fusion layer; analyzing IoT data in real time through an intelligent decision-making layer to create a pre-diagnosis work order, or manually creating a fault work order and submitting it to the intelligent decision-making layer; once the work order enters the system, the intelligent decision-making layer calculates a comprehensive scheduling score for each eligible maintenance worker, automatically assigns the work order to the worker with the highest score, and pushes all information to their communication terminal; simultaneously with work assignment, the resource coordination layer automatically checks warehouse inventory; after maintenance is completed, the results and time taken are recorded back to the system, forming a self-learning closed loop. The shortcomings of the existing technology are: relying on matching results with a preset fault model to determine whether equipment has malfunctioned leads to a high false alarm rate, resulting in delayed equipment maintenance response and inefficient allocation of maintenance resources. Summary of the Invention

[0003] In view of this, embodiments of the present invention provide a maintenance management method, device, equipment and medium for fault work orders, in order to solve the technical problem that judging whether a device has failed based on the matching result with a preset fault model results in a high fault misjudgment rate, which leads to a lag in the response of equipment maintenance and inefficient allocation of maintenance resources.

[0004] Firstly, a maintenance management method for fault work orders is provided, the method comprising: Obtain the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data of N candidate engineers, where N is a positive integer greater than 2. The data of the candidate engineers includes the candidate engineer's geographical coordinates, historical repair success rate, current load, and skill tags. The repair status of the faulty server device is determined based on its serial number. The repair status includes being within the warranty period, past the warranty period, and extended warranty period. When the faulty server equipment is under warranty or under extended warranty, the deviation between the real-time operating status and the historical operating status of the faulty server equipment is determined based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; the semantic similarity between the current fault and the historical fault of the faulty server equipment is determined based on the current fault type description text and the historical fault description text corresponding to each fault type. The state deviation and semantic similarity are weighted and summed to obtain the fault type evaluation index of the faulty server device. Based on the fault type evaluation index, the final fault type of the faulty server device is determined, and based on the final fault type, the adaptation and maintenance plan of the faulty server device is determined. The final fault type is matched with the data of N candidate engineers. Based on the matching results, the target engineer is determined. The fault work order is dispatched to the terminal device bound to the target engineer, and the adapted maintenance plan is sent.

[0005] Secondly, a maintenance management device for fault work orders is provided, the device comprising: The data acquisition module is used to acquire the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data of N candidate engineers, where N is a positive integer greater than 2. The data of the candidate engineers includes the candidate engineer's geographical coordinates, historical repair success rate, current load, and skill tags. The equipment maintenance status determination module is used to determine the maintenance status of the faulty server equipment based on the faulty server equipment number. The maintenance status includes within the warranty period, beyond the warranty period, and extended warranty period. The state deviation and semantic similarity determination module is used to determine the state deviation between the real-time operating state and the historical operating state of the faulty server equipment when the maintenance status of the faulty server equipment is within the warranty period or extended warranty period, based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; and to determine the semantic similarity between the current fault and the historical fault of the faulty server equipment based on the current fault type description text and the historical fault description text corresponding to each fault type. The adaptation and maintenance scheme determination module is used to perform weighted summation processing on the state deviation degree and the semantic similarity to obtain the fault type evaluation index of the faulty server device, determine the final fault type of the faulty server device based on the fault type evaluation index, and determine the adaptation and maintenance scheme of the faulty server device based on the final fault type. The target engineer identification and fault work order dispatch module is used to match the final fault type with the data of N candidate engineers, determine the target engineer based on the matching result, dispatch the fault work order to the terminal device bound to the target engineer, and send the adapted maintenance plan.

[0006] Thirdly, an electronic device is provided, the electronic device including a memory and a processor, wherein... Memory, used to store computer programs; The processor is used to execute programs stored in memory to implement the maintenance management method for fault work orders as described in the first aspect.

[0007] Fourthly, a computer-readable storage medium is provided, wherein a computer program is stored therein, and when the computer program is executed by a processor, it implements the maintenance management method for fault work orders as described in the first aspect.

[0008] The advantages of this invention compared to the prior art are: By determining the repair status of faulty server equipment, only equipment under warranty and those with extended warranty are allowed to create work orders, while the creation of work orders for out-of-warranty equipment is restricted, thus avoiding invalid work orders and unnecessary repair costs and reducing enterprise operation and maintenance costs. A dual-channel fault determination method, which weights and fuses the deviation of equipment operating parameters from their status and the semantic similarity of the fault description text, solves the problem of high misjudgment rates in single-dimensional fault judgment and accurately identifies the fault type. Based on the fault type, a suitable repair plan is determined, improving repair efficiency. Matching fault types with candidate engineer data enables refined allocation of human resources, reducing idle or overloaded human resources. Fault work orders and suitable repair plans are sent together to the terminal device bound to the engineer, shortening repair time and improving customer experience. Attached Figure Description

[0009] To more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings used in the description of the embodiments or the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0010] Figure 1This is a schematic diagram of an application environment for a maintenance management method for fault work orders provided in Embodiment 1 of the present invention; Figure 2 This is a schematic diagram illustrating the application environment of a maintenance management method for fault work orders provided in Embodiment 1 of the present invention. Figure 3 This is a flowchart illustrating a maintenance management method for fault work orders provided in Embodiment 1 of the present invention; Figure 4 This is a schematic diagram of the maintenance status verification logic of a maintenance management method for fault work orders provided in Embodiment 1 of the present invention; Figure 5 This is a flowchart of the fault work order flow in a fault work order maintenance management method provided in Embodiment 1 of the present invention. Figure 6 This is a schematic diagram of the structure of a maintenance management device for fault work orders provided in Embodiment 8 of the present invention; Figure 7 This is a schematic diagram of the structure of an electronic device provided in Embodiment 9 of the present invention. Detailed Implementation

[0011] To make the technical problems, technical solutions, and beneficial effects solved by this application clearer, the following detailed description is provided in conjunction with embodiments. It should be understood that the specific embodiments described herein are merely illustrative and not intended to limit the scope of this application.

[0012] The maintenance management method for fault work orders provided in Embodiment 1 of this invention can be applied to applications such as... Figure 1In the illustrated application environment, the user terminal communicates with the controller, the controller communicates with the server, and the controller communicates with the engineer. The user terminal includes, but is not limited to, handheld computers, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, cloud terminal devices, and personal digital assistants (PDAs). The controller can be implemented using a standalone controller or a controller cluster. The server device includes, but is not limited to, cloud server devices, edge computing server devices, storage server devices, database server devices, rack server devices, and blade server devices. The engineer's terminal includes, but is not limited to, handheld computers, desktop computers, laptops, ultra-mobile personal computers (UMPCs), netbooks, cloud terminal devices, and personal digital assistants (PDAs).

[0013] Specifically, such as Figure 2 As shown, the front-end module deployed on the user terminal device includes: Homepage module: Provides a search function. Users can retrieve relevant information about the faulty server device (such as warranty status data, device model, and historical repair records) by entering the device's serial number.

[0014] Fault Reporting Center: Provides workflow management functions for fault work orders. Users can create fault work orders (fill in fault description text, upload pictures and text document attachments, etc.), close fault work orders (close the order after acceptance), and check the current progress of fault work orders through the fault reporting center.

[0015] Information Operations and Maintenance Center: Provides visual display and message receiving functions. It displays the current node of fault work orders in real time (e.g., pending assignment, under repair, or under testing), estimated completion time, and engineer's processing progress; at the same time, it receives fault type determination results and maintenance plan recommendations pushed by the controller, and presents them to the user in card or list format.

[0016] The controller uses a built-in processor to implement the fault work order maintenance management method of this invention; the controller also has a backend module, including: Application Management Module: Primarily provides administrators with full-process management capabilities for fault work orders through the built-in operation and maintenance center. It supports filtering fault work orders by status (such as in progress, paused, completed, abnormal, and voided), and allows real-time viewing of work order progress, as well as operations such as enabling, deleting, or expediting.

[0017] Portal Management Module: Supports configuration of homepage module display content. Administrators can select a carousel or fixed image as the header of the front-end homepage module through the backend interface, and can add components such as query interfaces, FAQ interfaces, or big data statistics dashboards to adjust the display effect of the front-end homepage module as needed.

[0018] Organization Management Module: Supports the creation of basic roles such as users, administrators, and engineers, and assigns different operation permissions to each role: users can only create fault work orders, query the progress of fault work orders, and close fault work orders; administrators are responsible for the assignment, process management, and system configuration of fault work orders; engineers are responsible for fault repair and repair process recording.

[0019] System management module: Supports setting up messaging channels such as email, DingTalk, or WeChat Work. The controller system automatically sends reminders to the terminal devices of users, administrators, and engineers at key nodes such as fault work order creation, assignment, repair completion, and fault warning, ensuring timely information synchronization.

[0020] System Business Integration Module: Supports plugin content editing and template addition. When the maintenance business of faulty server equipment expands (e.g., adding GPU repair or hard drive data recovery services), administrators can use this module to flexibly add functional modules or adjust business logic to adapt to maintenance needs of different scales and scenarios.

[0021] See Figure 3 This is a flowchart illustrating a maintenance management method for fault work orders provided in Embodiment 1 of the present invention. Taking the application of this method on the controller side as an example, the maintenance management method for fault work orders may include the following steps: S100: Obtain the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data of N candidate engineers, where N is a positive integer greater than 2. The data of the candidate engineers includes the candidate engineer's geographical coordinates, historical repair success rate, current load, and skill tags. The faulty server device ID refers to the unique serial number of a faulty server device, used to retrieve information such as the server device's model, warranty status, historical repair records, and purchase date from the device management platform. Real-time operating parameters of the faulty server device refer to the operating data collected by the server device under its current operating conditions at the moment of the fault or in a short period before the fault occurred, including core temperature (e.g., GPU temperature), operating voltage, operating current, current power consumption, fan speed, hardware operation logs, and error parameters. Historical operating parameters corresponding to each fault type refer to the collection of operating parameters recorded for the same device or device of the same model when a confirmed fault (e.g., BIOS failure, power module damage, or GPU memory error) occurred in the historical fault ticket database. The user's current fault type description text refers to the description of the fault phenomenon (e.g., server cannot power on or power indicator light is off) filled in by the user (e.g., data center maintenance personnel or customers) when creating a fault ticket, using natural language. Historical fault description text corresponding to each fault type refers to the collection of all fault description texts belonging to the same fault type that have been archived in the historical fault ticket database. The candidate engineer data refers to the relevant data of potential repair personnel who meet the basic fault repair conditions and are currently available for work orders. This includes the candidate engineer's geographic coordinates, historical repair success rate, current workload, and skill tags. The candidate engineer's geographic coordinates refer to the latitude and longitude coordinates of their current location, obtained in real-time from the terminal device (e.g., mobile GPS) linked to the candidate engineer. The candidate engineer's historical repair success rate refers to the ratio of the number of fault work orders successfully repaired (without rework) to the total number of work orders completed by the candidate engineer within a past period (e.g., the last 3 months). The candidate engineer's current workload refers to the number of fault work orders currently being processed by the candidate engineer that have not yet been completed. The candidate engineer's skill tags refer to the classification tags of their repair technical capabilities extracted from their records or training certifications (e.g., H800 chip repair, server equipment BIOS debugging, or power supply fault repair).

[0022] In this embodiment, the current workload of the candidate engineer can be obtained by real-time statistics of the number of unfinished work orders of the candidate engineer, or by combining the average processing time of historical fault work orders corresponding to each fault type and converting it into the remaining available working hours of the candidate engineer.

[0023] S200, determine the repair status of the faulty server device according to the device number, the repair status including within the warranty period, beyond the warranty period, and extended warranty period; The maintenance status of a faulty server device refers to the different maintenance stages of the server device retrieved from the device management platform based on its serial number. These stages include: within the warranty period (the device displays a normal color and supports creating fault tickets), past the warranty period (the device displays gray and is not selectable), and extended warranty period (the device displays a special identifier and allows creating fault tickets according to extended warranty rules). This status is used to regulate the creation of fault tickets for the faulty server device. Within the warranty period means the current time of the faulty server device is within the standard warranty period specified by the original manufacturer. Past the warranty period means the current time of the faulty server device has exceeded the standard warranty period specified by the original manufacturer. Extended warranty period refers to a paid extended warranty service contract separately signed by the user before or after the original manufacturer's warranty expires, and the current time is within the valid warranty period specified in the extended warranty service contract.

[0024] In this embodiment, as Figure 3 As shown, when the faulty server equipment is within the warranty period and there is no human-caused damage exceeding the warranty terms, users can submit fault tickets normally and enjoy the original manufacturer's free maintenance services. Fault ticket creation is supported free of charge, and the subsequent fault diagnosis and dispatch process can proceed normally. When the faulty server equipment is beyond the warranty period and no valid extended warranty contract has been signed, free fault ticket creation is not supported. If repair is required, a separate self-paid repair process must be followed. When the faulty server equipment has an extended warranty period, users can initiate repair requests and create fault tickets normally, just like with faulty server equipment within the warranty period, and enjoy maintenance rights according to the terms of the extended warranty contract.

[0025] S300, when the maintenance status of the faulty server equipment is within the warranty period or extended warranty period, determine the state deviation between the real-time operating status and the historical operating status of the faulty server equipment based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; determine the semantic similarity between the current fault and the historical fault of the faulty server equipment based on the current fault type description text and the historical fault description text corresponding to each fault type. Among them, state deviation is a quantitative indicator used to measure the degree of difference between the current real-time operating parameters of a faulty server device and the historical operating parameter sets corresponding to each fault type. Semantic similarity is a quantitative indicator used to measure the degree of semantic similarity between the current fault description text input by the user and the historical fault description text sets corresponding to each fault type.

[0026] S400, perform weighted summation on the state deviation and the semantic similarity to obtain the fault type evaluation index of the faulty server device, determine the final fault type of the faulty server device based on the fault type evaluation index, and determine the adaptation and maintenance plan for the faulty server device based on the final fault type. Among them, the fault type evaluation index refers to a comprehensive quantitative indicator calculated for each candidate fault type to measure the credibility of the corresponding fault type in the current faulty server device. The final fault type refers to the fault type with the highest fault type evaluation index selected from all candidate fault types as the final fault type of the faulty server device. The adaptive repair plan refers to the optimal repair handling method that matches the determined final fault type, extracted from the device storage database or historical fault ticket database (e.g., prioritizing online remote repair for BIOS faults, prioritizing on-site pickup repair for hardware damage faults, etc.), including standardized repair steps, required parts, and repair tools.

[0027] In this embodiment, when performing weighted summation, the weights of state deviation and semantic similarity can be automatically fine-tuned using gradient descent.

[0028] S500: Match the final fault type with the data of N candidate engineers, determine the target engineer based on the matching result, dispatch the fault work order to the terminal device bound to the target engineer, and send the adapted maintenance plan.

[0029] The target engineer refers to the maintenance engineer with the highest overall suitability and best suitability to take on this fault work order, determined through multi-dimensional data matching between the final fault type and candidate engineers. A fault work order is an electronic task document recording a repair request for a faulty server device, including information such as the faulty server device's serial number, model, geographical coordinates, and final fault type. The terminal device linked to the target engineer refers to the mobile or fixed device used by the target engineer to receive fault work order notifications, view repair information, and provide operational feedback.

[0030] In this embodiment, as Figure 5 The diagram shown is a flowchart of the fault work order process, which mainly includes: User submission: Users create a fault work order through the front-end fault reporting center, fill in the fault information of the faulty server equipment and submit it; Administrator assignment: After receiving a fault work order, the controller system assigns the fault work order, and the administrator reviews and confirms it (emergency fault work orders can be automatically assigned). Engineer order acceptance and repair: After receiving a fault order, the engineer selects the appropriate repair solution and performs operations such as receiving, repairing and testing. User Acceptance and Clearance: After the engineer completes the repair and the test verification is qualified, the goods are shipped to the user. The clearance is completed after the user accepts the goods and there are no problems.

[0031] In this embodiment, for complex final fault types, the controller will automatically associate and generate an AR remote collaboration tool entry point when dispatching a fault work order, and push this entry point to the terminal device bound to the target engineer. When the target engineer encounters difficult problems at the repair site, he can click on the entry point through the bound terminal device to transmit the on-site fault scene (video stream or high-definition image) to the controller in real time. The controller system will automatically match the most suitable remote expert according to the final fault type. The remote expert will provide immediate technical support by adding annotations to the on-site fault scene or sharing the screen to display similar historical cases or test programs.

[0032] The fault work order maintenance management method in this embodiment determines the maintenance status of faulty server equipment, allowing only in-warranty and extended-warranty equipment to create work orders, while restricting the creation of fault work orders for out-of-warranty equipment. This avoids invalid work order assignments and unnecessary maintenance costs, reducing enterprise operation and maintenance costs. A weighted fusion of the deviation of equipment operating parameters from their status and the semantic similarity of the fault description text in a dual-channel fault determination process solves the problem of high misjudgment rates in single-dimensional fault judgment, accurately identifying the fault type. Based on the fault type, a suitable maintenance plan is determined, improving maintenance efficiency. Matching fault types with candidate engineer data enables refined allocation of human resources, reducing idle or overloaded human resources. The fault work order and the suitable maintenance plan are sent together to the engineer's bound terminal device, shortening maintenance time and improving customer experience.

[0033] In Example 2, the repair status of the faulty server device is determined based on its serial number. The repair status includes being within the warranty period, past the warranty period, and extended warranty period, including: When the faulty server equipment is under warranty, the model, usage environment data, historical maintenance frequency, and failure patterns of the same model of the faulty server equipment are obtained. Based on the usage environment data, historical maintenance frequency, and failure patterns of the same model of the faulty server equipment, a risk level assessment of the faulty server equipment is obtained. Based on the risk level assessment, high-frequency faulty equipment is marked, and a device risk warning is generated. The usage environment data of the faulty server equipment refers to the external environment (e.g., operating time under high temperature) or operating environment parameters (e.g., overclocking records) in which the faulty server equipment operates. The historical maintenance frequency of the faulty server equipment refers to the number of times a fault occurred and was repaired within a past period (e.g., the last 12 months) from the historical maintenance records of the server equipment, retrieved from the equipment management platform based on the faulty server equipment's serial number. The fault pattern of the same model of equipment refers to the fault distribution characteristics (e.g., high-frequency fault types, fault triggering factors, high-incidence periods, etc.) of all equipment of the same model as the currently faulty server equipment, retrieved from the historical fault work order database based on the faulty server equipment's model. The risk level assessment of the faulty server equipment is a quantitative indicator (e.g., high-risk, medium-risk, or low-risk) used to measure the potential probability of a fault occurring. High-frequency faulty equipment refers to server equipment whose risk level exceeds a preset risk level threshold and has a high probability of subsequent faults. Equipment risk warnings refer to automatically generated early warning information for high-frequency faulty equipment (e.g., equipment aging risk warnings).

[0034] In this embodiment, the risk level assessment can be obtained by weighted summation of the usage environment data of the faulty server equipment, historical maintenance frequency, and failure patterns of the same model of equipment.

[0035] When the faulty server equipment is out of warranty, the subsequent fault determination process is terminated and a message is displayed indicating that the fault work order cannot be created. In this embodiment, when the repair status of the faulty server equipment is beyond the warranty period, the subsequent steps related to fault diagnosis are immediately stopped, and a clear pop-up window or page text reminder is returned to the customer or administrator, informing them that the current faulty server equipment cannot create a free fault ticket because it is beyond the warranty period.

[0036] When the faulty server equipment is under extended warranty, determine the repair rights and remaining repair times for the faulty server equipment.

[0037] The repair rights for faulty server equipment refer to the scope and rules of maintenance services stipulated in the paid extended warranty service contract separately signed by the user during the extended warranty period. This includes service types (such as on-site repair, parts replacement, and remote inspection), covered components (such as specific components like motherboards and power supplies, or the entire machine), and cost allocation. The remaining repair attempts for the faulty server equipment refer to the remaining number of free repairs available as stipulated in the paid extended warranty service contract.

[0038] In this embodiment, the remaining number of repair attempts for the faulty server device is the difference between the maximum number of repair attempts stipulated in the paid extended warranty service contract and the number of repair attempts currently used.

[0039] The fault ticket maintenance management method in this embodiment differentiates the handling of faulty server equipment across three maintenance statuses: within the warranty period, past the warranty period, and extended warranty period. For equipment within the warranty period, a risk level assessment is performed based on usage environment data, historical maintenance frequency, and fault patterns of similar models. High-frequency faulty equipment is flagged and risk warnings are generated, shifting from passive response to proactive prevention. This helps users identify equipment aging risks in advance, reducing the probability of sudden and recurring failures and extending equipment lifespan. For equipment past the warranty period, the subsequent fault determination process is terminated, and a fault ticket cannot be created. This avoids useless fault identification and additional maintenance costs, saves computing resources, and effectively reduces maintenance costs. For equipment in the extended warranty period, maintenance rights and remaining maintenance attempts are determined, facilitating compliance verification before fault ticket creation. This avoids maintenance disputes caused by exceeding the allowed number of uses or improper maintenance, improving maintenance efficiency and user experience.

[0040] In Embodiment 3, when the faulty server equipment is under warranty or under extended warranty, the deviation between the real-time operating status and the historical operating status of the faulty server equipment is determined based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type, including: When the faulty server equipment is under warranty or under extended warranty, numerical feature extraction processing is performed on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type to obtain the real-time feature vector and historical fault vector of the faulty server equipment. Numerical feature extraction refers to the process of converting raw operating parameters into numerical vectors that can be used for distance calculation. The real-time feature vector of a faulty server device is a numerical vector representing the current real-time operating state of the faulty server device, obtained after extracting numerical features from its real-time operating parameters. Historical fault vectors are a collection of multiple numerical vectors obtained after extracting numerical features from historical operating parameters corresponding to various fault types.

[0041] The deviation between the real-time operating state and the historical operating state of the faulty server device is determined based on the Mahalanobis distance between the real-time feature vector and the historical fault vector.

[0042] Mahalanobis distance is a multidimensional spatial distance metric used to measure the deviation between real-time feature vectors and historical fault vector sets. It is mainly used in fields such as anomaly detection, fault diagnosis, classification, and outlier screening, and can eliminate interference caused by differences in the dimensions and correlations between different dimensions of data.

[0043] The fault work order maintenance management method in this embodiment extracts numerical features and calculates Mahalanobis distance for faulty server equipment within or after the warranty period. It maps the real-time operating parameters of the equipment with the historical operating parameters corresponding to each fault type to a unified multi-dimensional space, effectively eliminating the dimensional differences and correlation interference between parameters of different dimensions. This allows the calculated state deviation to more accurately reflect the true deviation between the current operating state and the historical fault operating state of the equipment, reducing the risk of fault misjudgment caused by fluctuations in a single parameter or coupling effects between parameters. It also provides data support for subsequent dual-channel weighted fusion for fault type judgment.

[0044] In Embodiment 4, the semantic similarity between the current fault and historical faults of the faulty server device is determined based on the current fault type description text and the historical fault description text corresponding to each fault type, including: Text feature extraction processing is performed on the current fault type description text and the historical fault description text corresponding to each fault type to obtain the real-time fault text vector and the historical fault text vector of the faulty server device. Text feature extraction refers to the process of converting unstructured natural language text into semantic vectors that can be used for similarity calculation. The real-time fault text vector of a faulty server device is a semantic vector representing the current fault of the server device, obtained by extracting text features from the user-input description of the current fault type. The historical fault text vector is a collection of semantic vectors obtained by extracting text features from historical fault descriptions corresponding to each fault type.

[0045] In this embodiment, the BERT pre-trained model is used to extract text features from the current fault type description text and the historical fault description text corresponding to each fault type, so as to obtain the real-time fault text vector and the historical fault text vector of the faulty server device.

[0046] Based on the real-time fault text vector and the historical fault text vector, the semantic similarity between the current fault type description text and each of the historical fault description texts is obtained.

[0047] In this embodiment, the cosine similarity between the real-time fault text vector and the historical fault text vector can be calculated as the semantic similarity between the current fault type description text and each historical fault description text.

[0048] The fault work order maintenance management method in this embodiment converts the unstructured current fault type description text and the historical fault description text corresponding to each fault type into semantic vectors and calculates the semantic similarity between the two. This effectively reduces the risk of fault misjudgment caused by different expression habits, description differences or ambiguity of the reporting personnel, and provides data support for subsequent dual-channel weighted fusion to determine the fault type.

[0049] In Embodiment 5, the steps for determining the adaptation and maintenance scheme include: Based on the final fault type, historical fault work orders are filtered from the historical fault work order database to obtain historical fault work orders that match the final fault type. From the historical fault work orders, the average processing time, implementation success rate, and customer satisfaction of each maintenance solution for each historical fault work order that matches the final fault type are extracted. The historical fault work order database is an archived database storing all completed historical fault work orders, including fault type, corresponding repair solutions, average processing time for each repair solution, implementation success rate, and customer satisfaction data. The average processing time for each repair solution is the arithmetic mean of the man-hours consumed by various repair methods for the final fault type (e.g., online remote repair, on-site pickup repair, a combination of online and on-site repair, and mail-in repair). The implementation success rate for each repair solution is the ratio of the number of successful repairs to the total number of repairs for various repair methods for the final fault type. The customer satisfaction rate for each repair solution is the arithmetic mean of the satisfaction ratings given by users after the actual fault repair is completed for various repair methods for the final fault type.

[0050] The average processing time, the success rate of implementation, and the customer satisfaction are weighted and summed to obtain a comprehensive evaluation value for each of the repair solutions. The repair solutions are then ranked according to the magnitude of the comprehensive evaluation value to determine the adaptation priority of each repair solution. The overall evaluation value is a quantitative assessment indicator used to measure the overall performance of each repair solution across three dimensions: average processing time, implementation success rate, and customer satisfaction. It reflects the effectiveness of each repair solution in addressing the final type of fault in the server equipment. Adaptation priority refers to the recommended order of repair solutions after ranking them according to the overall evaluation value; the higher the overall evaluation value, the higher the adaptation priority of the corresponding repair solution.

[0051] The repair scheme with the highest adaptation priority is selected as the adaptation repair scheme.

[0052] In this embodiment, the repair plan with the highest adaptation priority is used as the adaptation and repair plan for the faulty server equipment.

[0053] The fault work order maintenance management method in this embodiment performs a weighted quantitative evaluation of three dimensions—average processing time, implementation success rate, and customer satisfaction—for each maintenance solution matching the final fault type in the historical fault work order database. This comprehensive evaluation value is then ranked and prioritized, and the maintenance solution with the highest priority (comprehensive evaluation value) is selected as the appropriate maintenance solution. By selecting appropriate maintenance solutions based on massive amounts of real historical operation and maintenance data, the method avoids subjective biases and evaluation limitations caused by relying on personal experience or single indicators. It balances maintenance time, reliability, and user experience, achieving intelligent and accurate recommendation of maintenance solutions. This improves maintenance efficiency and quality, enhances customer satisfaction, and provides standardized operating guidelines for maintenance engineers, effectively reducing the trial-and-error costs of repeated on-site debugging.

[0054] In Example 6, the final fault type is matched with the data of N candidate engineers, and the target engineer is determined based on the matching results, including: Obtain the geographical coordinates of the faulty server device; Among them, geographic location coordinates refer to the location information of the faulty server device in physical space, which can be represented in the form of latitude and longitude coordinates or regional point coordinates.

[0055] Based on the geographical coordinates of the faulty server device and the geographical coordinates of the N candidate engineers, determine the geographical distance between the faulty server device and the N candidate engineers; Geographical distance is a quantitative indicator used to measure the spatial proximity between the faulty server equipment and each candidate engineer.

[0056] In this embodiment, the geographical distance refers to the spatial distance calculated based on the geographical coordinates of the faulty server device and the geographical coordinates of the candidate engineer, which is then further converted into the travel time corresponding to the optimal route using a map API.

[0057] The geographical distance, the historical repair success rate, and the current load are weighted and summed to obtain the matching degree between N candidate engineers and the fault work order; Among them, the matching degree refers to a comprehensive quantitative indicator used to evaluate the degree of fit between each candidate engineer and the current fault ticket in three dimensions: spatial proximity, historical reliability, and current availability.

[0058] In this embodiment, when performing weighted summation on geographical distance, historical repair success rate, and current load, the corresponding weighting coefficients are 10%, 60%, and 30%, respectively.

[0059] The target engineer is determined based on the matching degree and the skill tags of the N candidate engineers.

[0060] In this embodiment, based on the skill requirements of the final fault type, candidate engineers whose skill tags do not match are first excluded. Then, the matching degree between the remaining candidate engineers and the fault work order is sorted from largest to smallest, and the engineer with the highest matching degree is selected as the target engineer.

[0061] The fault work order maintenance management method in this embodiment first obtains the geographical coordinates of the faulty server equipment and each candidate engineer, and calculates the geographical distance between them (optimal route time). Then, it combines the historical repair success rate with the current load and performs a weighted sum to obtain the matching degree between the candidate engineer and the fault work order. Finally, it determines the target engineer based on skill tags and matching degree ranking. By comprehensively considering multiple factors such as response time, repair capability, workload, and professional skills, it effectively avoids the subjectivity of manual work order dispatch and the limitations of single-indicator decision-making, realizing multi-dimensional data-driven intelligent work order dispatch. This significantly shortens the fault response time, improves the success rate of repairs, optimizes human resource allocation, reduces the risk of engineer overload, and improves the overall operation and maintenance efficiency and customer satisfaction in fault handling.

[0062] In Embodiment Seven, the maintenance management method for the fault work order further includes: Obtain the current fault image of the faulty server device and the current text data of the faulty work order uploaded by each node during the workflow of the fault work order. In this context, each node in the workflow of a fault work order refers to a key stage within the entire lifecycle of the work order, from creation, assignment, repair, testing to closure. Current fault images refer to image data related to the faulty server equipment uploaded at each workflow node of the current fault work order (e.g., photos of burnt power connectors, abnormal indicator lights, or screens displaying error codes). Current text data refers to textual information recorded during the workflow of the current fault work order (e.g., user fault description attachments, engineer's repair process records, and test reports).

[0063] The current fault image is subjected to image recognition processing to obtain the feature vector of the current fault image; Image recognition processing refers to the process of analyzing fault images using big data image recognition technologies (such as convolutional neural networks (CNNs), image classification, object detection, or feature extraction models), automatically identifying key visual features in the images (such as burn marks on power connectors, the color of abnormal indicator lights, or numerical areas displaying error codes), and converting them into numerical vectors. The feature vector of the current fault image refers to the numerical vector obtained by converting the key visual features extracted from the current fault image through image recognition processing.

[0064] Keyword extraction is performed on the current text data to obtain the keywords of the fault work order; Keyword extraction processing refers to the process of using natural language processing techniques (such as TF-IDF, TextRank, word frequency statistics, or keyword extraction based on pre-trained models) to extract words or phrases from unstructured text data that represent the core content, fault characteristics, and repair points of the fault work order. The keywords of the current fault work order refer to words or phrases (such as power module, inability to power on, motherboard replacement, etc.) obtained through keyword extraction processing that summarize the core information of the current fault work order.

[0065] In this embodiment, TF-IDF text retrieval technology is used to extract keywords from the current text data of the current fault work order to obtain the keywords of the fault work order.

[0066] Based on the feature vector of the current fault image and the feature vector of the historical fault images in the historical fault work order database, the image similarity between the current fault image and the historical fault images is obtained. Image similarity is a quantitative indicator used to measure the degree of similarity between current fault images and historical fault images in key visual features.

[0067] In this embodiment, each historical fault image stored in the historical fault work order database undergoes the same image recognition processing when it is added to the database, and the feature vector of the corresponding historical fault image is extracted and saved.

[0068] In this embodiment, the cosine similarity between the feature vector of the current fault image and the feature vector of the historical fault images can be used as the image similarity between the current fault image and the historical fault images.

[0069] Based on the keywords of the current fault work order and the keywords of historical fault work orders in the historical fault work order database, the text similarity between the current fault work order and the historical fault work orders is obtained. Text similarity is a quantitative indicator used to measure the degree of similarity between the current fault work order and historical fault work orders in terms of text content.

[0070] In this embodiment, each historical fault work order stored in the historical fault work order database also undergoes the same keyword extraction process when it is added to the database, and the corresponding keywords of the historical fault work order are saved.

[0071] In this embodiment, the Jaccard similarity between the keywords of the current fault work order and the keywords of the historical fault work orders can be used as the text similarity between the current fault work order and the historical fault work orders.

[0072] The image similarity and text similarity are weighted and summed. Based on the weighted summation result, historical fault work orders similar to the fault work order are identified, and the troubleshooting steps and test parameters of the repair plan corresponding to the historical fault work order are sent to the terminal device bound to the target engineer.

[0073] In this embodiment, historical fault work orders similar to the current fault work order are selected by sorting the weighted sum of image similarity and text similarity, and the troubleshooting steps and test parameters of the corresponding repair solutions are sent to the terminal device bound to the target engineer.

[0074] The fault work order maintenance management method in this embodiment collects fault images and text data uploaded at each node during the current fault work order's workflow. It then uses image recognition processing to extract image feature vectors and natural language processing to extract text keywords. These are then matched against a historical fault work order database using both image and text similarity. Based on the weighted sum of these two results, the method accurately retrieves the most similar historical fault work order and automatically pushes its corresponding troubleshooting steps and test parameters to the terminal device bound to the target engineer. By matching similar historical fault work orders, it provides mature and standardized historical maintenance experience and technical parameter support for on-site maintenance work, significantly reducing the time engineers spend repeatedly troubleshooting and the cost of trial and error on-site. This improves the first-time repair success rate and overall maintenance efficiency, achieving standardization and intelligentization of maintenance services.

[0075] Embodiment 8 of this application also provides a maintenance management device 60 for fault work orders. Please refer to... Figure 6 ,include: The data acquisition module is used to acquire the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data of N candidate engineers, where N is a positive integer greater than 2. The data of the candidate engineers includes the candidate engineer's geographical coordinates, historical repair success rate, current load, and skill tags. The equipment maintenance status determination module is used to determine the maintenance status of the faulty server equipment based on the faulty server equipment number. The maintenance status includes within the warranty period, beyond the warranty period, and extended warranty period. The state deviation and semantic similarity determination module is used to determine the state deviation between the real-time operating state and the historical operating state of the faulty server equipment when the maintenance status of the faulty server equipment is within the warranty period or extended warranty period, based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; and to determine the semantic similarity between the current fault and the historical fault of the faulty server equipment based on the current fault type description text and the historical fault description text corresponding to each fault type. The adaptation and maintenance scheme determination module is used to perform weighted summation processing on the state deviation degree and the semantic similarity to obtain the fault type evaluation index of the faulty server device, determine the final fault type of the faulty server device based on the fault type evaluation index, and determine the adaptation and maintenance scheme of the faulty server device based on the final fault type. The target engineer identification and fault work order dispatch module is used to match the final fault type with the data of N candidate engineers, determine the target engineer based on the matching result, dispatch the fault work order to the terminal device bound to the target engineer, and send the adapted maintenance plan.

[0076] Optionally, the above-mentioned equipment maintenance status determination module includes: The equipment processing module during the warranty period is used to obtain the model, usage environment data, historical maintenance frequency and failure patterns of the same model of the faulty server equipment when the maintenance status of the faulty server equipment is within the warranty period, and to obtain a risk level assessment of the faulty server equipment based on the usage environment data, the historical maintenance frequency and the failure patterns of the same model of the faulty server equipment, and to mark high-frequency faulty equipment and generate equipment risk warnings based on the risk level assessment. The out-of-warranty equipment processing module is used to terminate the subsequent fault determination process and prompt that the fault work order cannot be created when the repair status of the faulty server equipment is out of warranty. The extended warranty equipment processing module is used to determine the repair rights and remaining repair times of the faulty server equipment when the repair status of the faulty server equipment is under extended warranty.

[0077] Optionally, the aforementioned state deviation and semantic similarity determination module includes: The numerical feature extraction and processing module is used to perform numerical feature extraction processing on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type when the maintenance status of the faulty server equipment is within the warranty period or the extended warranty period, so as to obtain the real-time feature vector and the historical fault vector of the faulty server equipment. The state deviation calculation module is used to determine the state deviation between the real-time operating state and the historical operating state of the faulty server device based on the Mahalanobis distance between the real-time feature vector and the historical fault vector.

[0078] Optionally, the aforementioned state deviation and semantic similarity determination module further includes: The text feature extraction and processing module is used to perform text feature extraction processing on the current fault type description text and the historical fault description text corresponding to each fault type, respectively, to obtain the real-time fault text vector and the historical fault text vector of the fault server device. The semantic similarity calculation module is used to obtain the semantic similarity between the current fault type description text and each of the historical fault description texts based on the real-time fault text vector and the historical fault text vector.

[0079] Optionally, the above-mentioned adaptation and maintenance solution determination module includes: The historical fault work order filtering module is used to filter from the historical fault work order library according to the final fault type, obtain historical fault work orders that are consistent with the final fault type, and extract the average processing time, implementation success rate and customer satisfaction of each maintenance solution of the historical fault work orders that are consistent with the final fault type from the historical fault work orders. The adaptation priority determination module is used to perform weighted summation on the average processing time, the implementation success rate and the customer satisfaction to obtain a comprehensive evaluation value for each of the repair solutions, and to determine the adaptation priority of each of the repair solutions according to the size of the comprehensive evaluation value. The adaptation and maintenance scheme selection module is used to select the maintenance scheme with the highest adaptation priority as the adaptation and maintenance scheme.

[0080] Optionally, the aforementioned target engineer identification and fault ticket dispatch module includes: A geographic location coordinate acquisition module is used to acquire the geographic location coordinates of the faulty server device; The geographic distance determination module is used to determine the geographic distance between the faulty server device and the N candidate engineers based on the geographic coordinates of the faulty server device and the geographic coordinates of the N candidate engineers. The matching degree calculation module is used to perform weighted summation on the geographical distance, the historical repair success rate and the current load to obtain the matching degree between N candidate engineers and the fault work order; The target engineer matching module is used to determine the target engineer based on the matching degree and the skill tags of the N candidate engineers.

[0081] Optionally, the maintenance management device for the aforementioned fault work orders also includes: The fault data acquisition module is used to acquire the current fault image of the faulty server device and the current text data of the faulty work order uploaded by each node during the process of the fault work order. The image recognition processing module is used to perform image recognition processing on the current fault image to obtain the feature vector of the current fault image; The keyword extraction and processing module is used to extract keywords from the current text data to obtain the keywords of the current fault work order; The image similarity acquisition module is used to obtain the image similarity between the current fault image and the historical fault images in the historical fault work order database based on the feature vector of the current fault image and the feature vector of the historical fault images in the historical fault work order database. The text similarity acquisition module is used to obtain the text similarity between the current fault work order and the historical fault work orders in the historical fault work order database based on the keywords of the current fault work order and the keywords of the historical fault work orders. The similar historical fault work order determination module is used to perform weighted summation processing on the image similarity and the text similarity, and based on the weighted summation result, determine the historical fault work orders similar to the fault work order, and send the troubleshooting steps and test parameters of the repair plan corresponding to the historical fault work order to the terminal device bound to the target engineer.

[0082] Specific limitations regarding the maintenance management device for fault work orders can be found in the limitations of the maintenance management method for fault work orders mentioned above, and will not be repeated here. Each module in the aforementioned maintenance management device for fault work orders can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in the processor of the electronic device in hardware form or independently of it, or stored in the memory of the electronic device in software form, so that the processor can call and execute the corresponding operations of each module.

[0083] Embodiment 9 of this application also provides an electronic device 70, please refer to... Figure 7It includes a memory and a processor, wherein the memory is used to store computer programs; the processor is used to execute the programs stored in the memory to implement the fault work order maintenance management method described in any embodiment of this application.

[0084] Embodiment 10 of this application also provides a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the fault work order maintenance management method described in any embodiment of this application.

[0085] In this application, "multiple" refers to two or more.

[0086] In this application, unless otherwise expressly defined, the terms "installation," "connection," and "linking" should be interpreted broadly. For example, they can refer to a fixed connection, a detachable connection, or an integral connection; they can refer to a mechanical connection or an electrical connection; they can refer to a direct connection or an indirect connection through an intermediate medium; and they can refer to the internal connection between two components. Those skilled in the art can understand the specific meaning of the above terms in this application based on the specific circumstances.

[0087] The terms “first,” “second,” “third,” “fourth,” etc., in this application (if present) are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence.

[0088] In this application, the term "and / or" is merely a description of the relationship between related objects, indicating that three relationships can exist. For example, A and / or B can represent: A existing alone, A and B existing simultaneously, or B existing alone. Additionally, in this application, the character " / " generally indicates that the preceding and following related objects have an "or" relationship.

[0089] Unless otherwise specified, all steps in this application may be performed sequentially or randomly. For example, if the method includes steps A and B, it means that the method may include steps A and B performed sequentially, or it may include steps B and A performed sequentially. For example, if the method may also include step C, it means that step C may be added to the method in any order. For example, the method may include steps A, B, and C, or it may include steps A, C, and B, or it may include steps C, A, and B, etc.

[0090] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A maintenance management method for fault work orders, characterized in that, The maintenance management method for the fault work orders includes: The system obtains the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data for N candidate engineers, where N is a positive integer greater than 1. The data for the candidate engineers includes their geographical coordinates, historical repair success rate, current load, and skill tags. The repair status of the faulty server device is determined based on its serial number. The repair status includes being within the warranty period, past the warranty period, and extended warranty period. When the faulty server equipment is under warranty or under extended warranty, the deviation between the real-time operating status and the historical operating status of the faulty server equipment is determined based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; the semantic similarity between the current fault and the historical fault of the faulty server equipment is determined based on the current fault type description text and the historical fault description text corresponding to each fault type. The state deviation and semantic similarity are weighted and summed to obtain the fault type evaluation index of the faulty server device. Based on the fault type evaluation index, the final fault type of the faulty server device is determined, and based on the final fault type, the adaptation and maintenance plan of the faulty server device is determined. The final fault type is matched with the data of N candidate engineers. Based on the matching results, the target engineer is determined. The fault work order is dispatched to the terminal device bound to the target engineer, and the adapted maintenance plan is sent.

2. The maintenance management method for fault work orders according to claim 1, characterized in that, Based on the serial number of the faulty server device, the repair status of the faulty server device is determined. The repair status includes being within the warranty period, past the warranty period, and extended warranty period, including: When the faulty server equipment is under warranty, the model, usage environment data, historical maintenance frequency, and failure patterns of the same model of the faulty server equipment are obtained. Based on the usage environment data, historical maintenance frequency, and failure patterns of the same model of the faulty server equipment, a risk level assessment of the faulty server equipment is obtained. Based on the risk level assessment, high-frequency faulty equipment is marked, and a device risk warning is generated. When the faulty server equipment is out of warranty, the subsequent fault determination process is terminated and a message is displayed indicating that the fault work order cannot be created. When the faulty server equipment is under extended warranty, determine the repair rights and remaining repair times for the faulty server equipment.

3. The maintenance management method for fault work orders according to claim 1, characterized in that, When the faulty server equipment is under warranty or under extended warranty, the deviation between the real-time and historical operating states of the faulty server equipment is determined based on the real-time operating parameters and the historical operating parameters corresponding to each fault type, including: When the faulty server equipment is under warranty or under extended warranty, numerical feature extraction processing is performed on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type to obtain the real-time feature vector and historical fault vector of the faulty server equipment. The deviation between the real-time operating state and the historical operating state of the faulty server device is determined based on the Mahalanobis distance between the real-time feature vector and the historical fault vector.

4. The maintenance management method for fault work orders according to claim 1, characterized in that, Based on the current fault type description text and the historical fault description texts corresponding to each fault type, determine the semantic similarity between the current fault and historical faults of the faulty server device, including: Text feature extraction processing is performed on the current fault type description text and the historical fault description text corresponding to each fault type to obtain the real-time fault text vector and the historical fault text vector of the faulty server device. Based on the real-time fault text vector and the historical fault text vector, the semantic similarity between the current fault type description text and each of the historical fault description texts is obtained.

5. The maintenance management method for fault work orders according to claim 1, characterized in that, The steps for determining the adaptation and repair solution include: Based on the final fault type, filter from the historical fault work order database to obtain historical fault work orders that match the final fault type. Extract from the historical fault work orders that match the final fault type the average processing time, implementation success rate and customer satisfaction of each maintenance solution for each historical fault work order. The average processing time, the implementation success rate, and the customer satisfaction are weighted and summed to obtain a comprehensive evaluation value for each of the repair solutions. The repair solutions are then ranked according to the magnitude of the comprehensive evaluation value to determine the adaptation priority of each repair solution. The repair scheme with the highest adaptation priority is selected as the adaptation repair scheme.

6. The maintenance management method for fault work orders according to claim 1, characterized in that, The final fault type is matched with the data of N candidate engineers, and the target engineer is determined based on the matching results, including: Obtain the geographical coordinates of the faulty server device; Based on the geographical coordinates of the faulty server device and the geographical coordinates of the N candidate engineers, determine the geographical distance between the faulty server device and the N candidate engineers; The geographical distance, the historical repair success rate, and the current load are weighted and summed to obtain the matching degree between N candidate engineers and the fault work order; The target engineer is determined based on the matching degree and the skill tags of the N candidate engineers.

7. The maintenance management method for fault work orders according to claim 1, characterized in that, The maintenance management method for the fault work orders also includes: Obtain the current fault image of the faulty server device and the current text data of the faulty work order uploaded by each node during the workflow of the fault work order. The current fault image is subjected to image recognition processing to obtain the feature vector of the current fault image; Keyword extraction is performed on the current text data to obtain the keywords of the current fault work order; Based on the feature vector of the current fault image and the feature vector of the historical fault images in the historical fault work order database, the image similarity between the current fault image and the historical fault images is obtained. Based on the keywords of the current fault work order and the keywords of historical fault work orders in the historical fault work order database, the text similarity between the current fault work order and the historical fault work orders is obtained. The image similarity and text similarity are weighted and summed. Based on the weighted summation result, historical fault work orders similar to the fault work order are identified, and the troubleshooting steps and test parameters of the repair plan corresponding to the historical fault work order are sent to the terminal device bound to the target engineer.

8. A maintenance management device for fault work orders, characterized in that, The maintenance management device for the fault work order includes: The data acquisition module is used to acquire the faulty server device's ID, real-time operating parameters, historical operating parameters corresponding to each fault type, the user's current fault type description text, historical fault description text corresponding to each fault type, and data of N candidate engineers, where N is a positive integer greater than 2. The data of the candidate engineers includes the candidate engineer's geographical coordinates, historical repair success rate, current load, and skill tags. The equipment maintenance status determination module is used to determine the maintenance status of the faulty server equipment based on the faulty server equipment number. The maintenance status includes within the warranty period, beyond the warranty period, and extended warranty period. The state deviation and semantic similarity determination module is used to determine the state deviation between the real-time operating state and the historical operating state of the faulty server equipment when the maintenance status of the faulty server equipment is within the warranty period or extended warranty period, based on the real-time operating parameters of the faulty server equipment and the historical operating parameters corresponding to each fault type; and to determine the semantic similarity between the current fault and the historical fault of the faulty server equipment based on the current fault type description text and the historical fault description text corresponding to each fault type. The adaptation and maintenance scheme determination module is used to perform weighted summation processing on the state deviation degree and the semantic similarity to obtain the fault type evaluation index of the faulty server device, determine the final fault type of the faulty server device based on the fault type evaluation index, and determine the adaptation and maintenance scheme of the faulty server device based on the final fault type. The target engineer identification and fault work order dispatch module is used to match the final fault type with the data of N candidate engineers, determine the target engineer based on the matching result, dispatch the fault work order to the terminal device bound to the target engineer, and send the adapted maintenance plan.

9. An electronic device, characterized in that, The electronic device includes a memory and a processor, wherein, Memory, used to store computer programs; A processor is used to execute a program stored in a memory to implement the maintenance management method for fault work orders as described in any one of claims 1-7.

10. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores a computer program, which, when executed by a processor, implements the maintenance management method for fault work orders as described in any one of claims 1-7.

Images