Mobile operation and maintenance intelligent system based on large model application

Abstract

The invention discloses a mobile operation and maintenance intelligent system based on large model application, and the system comprises a positioning database which is used for obtaining the position information of equipment and operation and maintenance personnel in real time; the operation and maintenance personnel configuration management database is used for storing skill, performance and load data of operation and maintenance personnel; the equipment configuration management database is used for recording equipment states and maintaining historical data; the large model operation center module is used for converting customer voice into characters, performing semantic analysis and sending customer problems and demands; the order dispatching intelligent operation center module is used for calculating orders dispatched by the first n operation and maintenance personnel with higher scores based on an intelligent order dispatching calculation model according to the work orders, and starting an order grabbing mechanism; and the operation and maintenance personnel mobile terminal is used for acquiring operation and maintenance personnel positioning information in real time, updating the information to the positioning database and supporting operation and maintenance personnel order grabbing operation and work order state real-time updating. The whole-process intelligent management of the operation and maintenance work order is realized, and the customer service efficiency, the order sending accuracy and the equipment maintenance initiative are remarkably improved.

Description

Technical Field

[0001] The present invention relates to the field of information technology operation and maintenance management, and more specifically to a mobile operation and maintenance intelligent system based on large model applications. It involves the integration of large language models, intelligent algorithms and mobile Internet technologies, and is a full-process optimization solution that realizes intelligent customer service, precise work order dispatch and predictive equipment maintenance. Background Art

[0002] With the continuous development of information technology and the acceleration of digitalization, the daily operations of enterprises and institutions are highly dependent on stable and efficient IT infrastructure. Information operations and maintenance, as the core support for seamless integration of terminal devices, network environments, and business systems, not only covers back-end systems such as servers and databases, but also extends to the full lifecycle management of user-side infrastructure such as displays, network equipment, keyboard and mouse peripherals, and computer hardware and software.

[0003] In office scenarios, seemingly minor faults such as black screens, network interruptions, and peripheral failures may directly affect work efficiency. Computer system installation and configuration, operating system updates and maintenance, business data backup and recovery, and system page function debugging are key nodes related to business continuity and data security.

[0004] The existing information operation and maintenance technology solution is equipped with a customer service system. The customer service staff will answer the customer's fault reporting calls, place orders in the system, and then send the work orders to the operation and maintenance engineers for processing. After the operation and maintenance engineers complete the processing, they will feedback to the customer service staff for closed-loop management of the work orders.

[0005] When a customer service representative receives a problem report, they enter a work order into the system and then notify an operations engineer to handle it. Upon receiving the notification, the operations engineer logs into the system to review and print the work order, then contacts the customer and goes to the site to handle the work order. This workflow presents several challenges: First, customer service staff lack effective tools to efficiently answer customer questions and guide them through self-resolution. Second, they must manually fill out work orders, which leaves much room for improvement. Third, when dispatching work orders, the customer service representative is unaware of the operations engineer closest to the problem location, causing them to waste time traveling back and forth between locations, reducing operational efficiency. Fourth, if an operations engineer is currently handling a work order and receives the next one, they must return to the office to print the work order, which increases their time and reduces efficiency. Fifth, operations engineers passively accept work orders, failing to leverage their strengths and expertise, nor effectively exercising their initiative. Sixth, there is a lack of tools that can intelligently predict equipment health. Seventh, there is a lack of tools that can predict and analyze operations personnel.

[0006] Therefore, those skilled in the art are in urgent need of solving the above technical problems. Summary of the Invention

[0007] In view of this, the present invention provides a mobile operation and maintenance intelligent system based on large model applications, which at least partially solves the above technical problems, realizes the full-process intelligent management of operation and maintenance work orders, and significantly improves customer service efficiency, order dispatch accuracy and equipment maintenance initiative.

[0008] In order to achieve the above object, the present invention adopts the following technical solutions:

[0009] An embodiment of the present invention provides a mobile operation and maintenance intelligent system based on a large model application, including:

[0010] Positioning database, used to obtain real-time location information of equipment and operation and maintenance personnel;

[0011] Operation and maintenance personnel configuration management database, used to store operation and maintenance personnel skills, performance and load data;

[0012] Equipment configuration management database, used to record equipment status and maintenance history data;

[0013] The large model operation center module is used to convert customer voice into text and perform semantic analysis, and send customer questions and needs;

[0014] The dispatch intelligent calculation center module calculates the dispatch order based on the work order and the top n operators with the highest scores based on the intelligent dispatch calculation model, and activates the order grabbing mechanism;

[0015] The operation and maintenance personnel's mobile terminal is used to obtain the operation and maintenance personnel's location information in real time and update it to the positioning database, as well as support the operation and maintenance personnel's order grabbing operation and real-time update of the work order status.

[0016] Furthermore, it also includes: an equipment intelligent predictive maintenance module, which predicts equipment failures through a predictive equipment detection algorithm and generates a maintenance work order, which is pushed to the dispatching intelligent operation center module;

[0017] Furthermore, it also includes: a data prediction and analysis module, which analyzes and predicts the behavior trends of operation and maintenance personnel based on the work enthusiasm prediction model.

[0018] Furthermore, it also includes:

[0019] The personnel information maintenance module is used to add, delete, and modify the basic identity information, contact information, department and team affiliation, authority management, account credentials, shift and duty information, skills and expertise, and training record information of the operation and maintenance personnel in the operation and maintenance personnel configuration management database;

[0020] The equipment information maintenance module is used to add, delete and modify the equipment usage time data, equipment maintenance time data and equipment parts replacement time data in the equipment configuration management database.

[0021] Furthermore, it also includes:

[0022] The call information processing module is used to receive and process customer phone information and forward customer voice data to the large model operation center module;

[0023] The intelligent customer service module is used to receive customer questions and needs sent by the large model operation center module, generate recommended solutions or automatically generate work orders, and display them to customer service personnel for online answers.

[0024] Furthermore, the intelligent dispatch calculation model formula in the dispatch intelligent operation center module is as follows:

[0025] S = α × skill matching + β × regional matching + γ × time efficiency coefficient + δ × evaluation coefficient + ∈ × order grabbing activity - ζ × current load

[0026] The weight distribution is α=30%, β=25%, γ=20%, δ=15%, ∈=10%, and ζ=10%.

[0027] Furthermore, the equipment intelligent predictive maintenance module includes:

[0028] The health index calculation unit integrates the equipment's service life, maintenance records, and sensor data to calculate the equipment's health index. The formula is as follows:

[0029] Ht=w1(1-t / T)+w2((Δt / τ)+w3((s / s0)

[0030] Where w1+w2+w3=1 is the weight parameter; t / T is the ratio of the equipment’s service life to its design life; Δt / τ is the ratio of the time since the last maintenance to the average maintenance interval; s / s0 is the ratio of the current sensor reading to the alarm threshold;

[0031] The maintenance time prediction unit uses an exponential decay model to predict maintenance time. The formula is as follows:

[0032] T w =(1-Ht) / k×T s ;T w Predict maintenance time for equipment, T s is the remaining design life of the equipment; k is the attenuation coefficient;

[0033] A work order automatic generation unit triggers the creation of an emergency work order when the health index is lower than a first threshold and the equipment predicted maintenance time is lower than a second threshold; or triggers the creation of a regular work order when the health index is within a preset value range;

[0034] The work order pushing unit pushes the created work order to the dispatching intelligent operation center module.

[0035] Furthermore, the work enthusiasm prediction model formula of the data prediction and analysis module is as follows:

[0036] Calculate the comprehensive positivity index At:

[0037] At=0.4×P recen +0.3×(1-W current / W max )+0.2×H score +0.1×S match

[0038] P recen The performance average of the past three months, normalized to 0-1; W current is the current number of work orders to be processed, W max is the individual's maximum load; H score Score health; S match The matching degree of work order skills is calculated by the cosine similarity between the expertise label and the work order requirements;

[0039] The comprehensive positive index At combined with the attenuation factor λ = e -0.1×D Get the corrected value J:

[0040] J=At×λ×(1+0.2×I incentive )

[0041] I incentive It represents whether there is a performance reward period, and takes the value of 0 or 1; the correction value J is used as the work enthusiasm prediction score.

[0042] It can be seen from the above technical solutions that compared with the prior art, the present invention has the following technical advantages:

[0043] The present invention integrates positioning data, large-model semantic analysis, intelligent dispatching algorithms, and equipment health predictions. Through the synergistic effect of various modules, it realizes the full-chain intelligence of the operation and maintenance process, significantly improves efficiency, accuracy, and cost-effectiveness, and optimizes human resource management and equipment maintenance initiative. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, the drawings described below are merely embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on the provided drawings without paying any creative work.

[0045] Figure 1 This is a structural diagram of the mobile operation and maintenance intelligent system based on large model application provided by the present invention.

[0046] Figure 2 This is a work order flow chart of the mobile operation and maintenance intelligent system based on large model application provided by the present invention. DETAILED DESCRIPTION

[0047] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0048] Reference Figure 1 As shown, the embodiment of the present invention discloses a mobile operation and maintenance intelligent system based on large model application, including:

[0049] 1. Positioning database 100, used to obtain real-time location information of equipment and maintenance personnel. Equipment positioning uses real-time reading of location information from built-in RFID tags, dynamically updated based on the equipment's movement status. Maintenance personnel location information is based on real-time upload of location data by the maintenance personnel's mobile terminal 400 (such as a smartphone or dedicated device) using the GPS module.

[0050] For example, during data synchronization during implementation, the location information of equipment and personnel is refreshed every 30 seconds to ensure the real-time distance calculation when dispatching orders.

[0051] 2. Operation and maintenance personnel configuration management library 200, used to store operation and maintenance personnel skills, performance and load data; connected to the personnel information maintenance module 101.

[0052] The personnel information maintenance module 101 is used to add, delete, and modify the basic identity information, contact information, department and team affiliation, authority management, account credentials, shift and duty information, skills (elementary / intermediate / advanced) and specialties, and training records of the operation and maintenance personnel in the operation and maintenance personnel configuration management database 200. For example, basic identity information includes name, employee number, ID number, position / role; contact information includes office phone number, mobile phone number, emergency contact, corporate email address, and internal communication account.

[0053] The operation and maintenance personnel configuration management library 200 may also include, for example, historical work order completion time, customer rating star rating, current work order load, maximum concurrent work order capacity (e.g., 15 orders / week), order grabbing activity (number of successfully grabbed orders in the past week), health score (based on leave frequency and physical examination data), etc. All of these are maintained by the personnel information maintenance module 101.

[0054] For example, after each work order is completed, the system automatically updates the performance data, load status and order grabbing records of the operation and maintenance personnel.

[0055] 3. Equipment configuration management database 201, used to record equipment status and maintenance history data; connected to the equipment information maintenance module 102.

[0056] The device information maintenance module 102 is used to add, delete, and modify the device usage time data, device maintenance time data, and device parts replacement time data calculated up to the current time in the device configuration management database 201 .

[0057] The device information maintenance module 102 can add, delete, and modify basic device information, hardware configuration information, software and license information, network information, procurement and warranty information, physical location and usage status, maintenance and repair records, and associated dependencies in the device configuration management database 201. Basic device information includes device name, type, brand, model, serial number (SN), asset number, etc.; hardware configuration information includes CPU model / number of cores, memory capacity / type, hard drive specifications, etc.

[0058] The device configuration management database 201 records, for example, the device name, model, serial number, purchase date, design life, maintenance cycle, sensor data (such as temperature and voltage), and historical fault records. It also records the health status index: a real-time device health index (Ht) calculated based on the device's age, maintenance intervals, and sensor readings.

[0059] 4. The large model operation center module 202 is used to convert customer voice into text and perform semantic analysis, and send customer questions and needs; it is connected to the call information processing module 103 and the intelligent customer service module 303 respectively.

[0060] The call information processing module 103 is used to receive and process customer telephone information and forward the customer voice data to the large model operation center module 202, which converts the voice into text and performs semantic analysis.

[0061] The intelligent customer service module 303 is used to receive customer questions and needs sent by the large model operation center module 202. The intelligent customer service recommends answers to the questions, and the customer service staff provide online answers. For questions and needs that cannot be answered online, the intelligent customer service automatically initiates a work order to be filled out, and the customer service staff confirms the order after review.

[0062] 1) Speech processing and semantic analysis:

[0063] The customer's fault reporting voice is transmitted to the large model operation center module 202 through the call information processing module 103, and the voice recognition model (such as Whisper) is used to convert the voice into text.

[0064] For example, the intent of text content can be recognized based on a pre-trained semantic large model (such as GPT-4) to extract keywords (such as "black screen" and "network interruption").

[0065] 2) Solution Recommendation:

[0066] The system matches the semantic analysis results with the knowledge base and recommends standardized solutions to customer service staff (such as "restart the router" or "check the network cable connection").

[0067] If the problem cannot be solved by the recommended solution, a work order will be automatically generated and filled with information such as the fault type, device ID, and fault location.

[0068] During specific implementation, the big model operation center module 202 converts voice content into text content, and uses the analytical capabilities of the semantic big model to deeply understand customer problems and needs, and sends customer problems and needs to the intelligent customer service module 303. The intelligent customer service recommends answers to the questions, and the customer service staff provides online answers. For questions and needs that cannot be answered online, the intelligent customer service module 303 automatically initiates a work order to be filled in, and the customer service staff confirms the order after review.

[0069] 5. The dispatch intelligent calculation center module 300 calculates the dispatch order for the top n operators with the highest scores based on the work order and uses the intelligent dispatch calculation model to dispatch the work order, and activates the order grabbing mechanism;

[0070] The dispatch intelligent computing center module 300 uses intelligent calculations to determine dispatch personnel. After a customer service representative places an order, the positioning database 100 sends the positioning data to the dispatch intelligent computing center module 300. The operations and maintenance personnel configuration management database 200 and the equipment configuration management database 201 also send data to the dispatch intelligent computing center module 300. Based on the intelligent dispatch calculation model, the module comprehensively considers factors such as positioning distance, equipment status, and personnel expertise to dispatch the order. For example, the module prioritizes dispatching orders to the top five engineers, reserving 10 minutes for the engineers to independently secure orders.

[0071] Intelligent dispatch calculation model:

[0072] (1) Core calculation formula

[0073] Engineer comprehensive score (S) = α × skill matching + β × region matching + γ × time efficiency coefficient + δ × evaluation coefficient + ε × order grabbing activity - ζ × current load

[0074] Final assignment target: Select the top 5 engineers with the highest overall score (S)

[0075] (2) Parameter definition and calculation method

[0076] (1) Skill matching (weight α = 30%)

[0077] Calculation method: Skill matching degree = number of matching skills / total number of skills required by the work order × level coefficient

[0078] Level coefficients: Elementary = 0.8, Intermediate = 1.0, Advanced = 1.2.

[0079] Skill Updates: Engineer skills and levels are updated in real time.

[0080] (2) Regional matching (weight β = 25%): The distance between two points is calculated based on satellite positioning. If the distance between the engineer and the fault reporting point is ≤ 1 km, 1 point is awarded; if the distance is 1-5 km, 0.5 point is awarded; and if the distance is > 5 km, 0 point is awarded.

[0081] Calculation method: 1 (same area), 0.5 (adjacent area), 0 (other);

[0082] Area definition: The distance between the two points is calculated based on the satellite positioning of the two points. If the distance between the engineer and the fault reporting point is within 1 km, it is defined as the same area and scored 1 point. If the distance is 1-5 km, it is defined as the adjacent area and scored 0.5 points. If the distance is more than 5 km, it is defined as other areas and scored 0 points.

[0083] The distance calculated by two-point satellite positioning is:

[0084] d=R*arccos(sin(lat1)*sin(lat2)+cos(lat1)*cos(lat2)*cos(lon2-lon1))

[0085] Where d is the ground distance between the two points, R is the radius of the Earth, lat1 and lat2 are the latitudes of the two points, and lon1 and lon2 are the longitudes of the two points.

[0086] (3) Time efficiency coefficient (weight γ = 20%), which is the inverse of the ratio of the engineer's historical average completion time to the industry benchmark time. If five consecutive orders are completed ahead of schedule, an additional 10% bonus will be given.

[0087] Calculation method: 1 / (historical average completion time / industry benchmark time);

[0088] Additional reward: 5 consecutive orders completed ahead of schedule × 1.1 coefficient;

[0089] Historical average completion times and industry benchmark times are calculated instantly based on the current time.

[0090] (4) Customer evaluation coefficient (weight δ = 15%), which is the rate of positive reviews minus the rate of negative reviews in the past 30 days, multiplied by the star coefficient (3 stars 0.8, 4 stars 1.0, 5 stars 1.2).

[0091] Calculation method: (positive review rate in the past 30 days - negative review rate) × star coefficient;

[0092] Star rating factor: 3 stars = 0.8, 4 stars = 1.0, 5 stars = 1.2;

[0093] The rate of positive reviews, negative reviews and star ratings are all calculated in real time at the current time point.

[0094] (5) Order grabbing activity (weight ε = 10%), which is the square root of the number of successful orders grabbed in the past week multiplied by the response speed coefficient (response within 5 minutes is 1.2, and response within 5-10 minutes is 0.8).

[0095] Calculation method:

[0096] Response speed coefficient: Golden 5-minute coefficient = 1.2, 5-10 minutes = 0.8

[0097] The number of orders grabbed is calculated based on the current time node.

[0098] (6) Current load penalty (weight ζ = 10%), the ratio of the number of ongoing work orders to the maximum concurrency multiplied by 0.5.

[0099] Calculation method: Number of ongoing work orders / maximum concurrent work order × 0.5.

[0100] Order grabbing mechanism:

[0101] The system pushes work order notifications to the top 5 engineers with the highest overall scores, reserving a 10-minute window for grabbing orders.

[0102] If multiple engineers compete for an order, priority will be given to the one with the highest score; if no one responds, it will be automatically assigned to the engineer with the highest score.

[0103] 6. The operation and maintenance personnel mobile terminal 400 is used to obtain the operation and maintenance personnel's location information in real time and update it to the location database 100, as well as support the operation and maintenance personnel's order grabbing operation and real-time update of the work order status.

[0104] 7. The equipment intelligent predictive maintenance module 301 predicts equipment failures through a predictive equipment detection algorithm and generates a maintenance work order, which is then pushed to the dispatch intelligent operation center module 300;

[0105] The data from the equipment configuration management database 201 is sent to the equipment intelligent predictive maintenance module 301 for calculation and intelligent prediction. The module predicts when the equipment needs maintenance, automatically initiates a maintenance work order based on the time, and sends it to the dispatch intelligent calculation center module 300 for intelligent dispatching.

[0106] The equipment intelligent predictive maintenance module 301 specifically includes the following units:

[0107] 1) Health Index Calculation Unit, which integrates the equipment usage years, maintenance records, and sensor data to calculate the equipment health index. The formula is as follows:

[0108] Ht = w1(1 - t / T) + w2((Δt / τ) + w3((s / s0)

[0109] Where, w1 + w2 + w3 = 1 is the weight parameter; t / T is the ratio of the equipment's used years to the design life; Δt / τ is the ratio of the time since the last maintenance to the average maintenance interval; s / s0 is the ratio of the current sensor reading to the alarm threshold; Ht ∈ [0, 1], and the smaller the value, the more maintenance is required.

[0110] 2) Maintenance Time Prediction Unit, which uses an exponential decay model to predict the maintenance time. The formula is as follows:

[0111] T w = (1 - Ht) / k × T s ; T w is the predicted maintenance time of the equipment, T s is the remaining design life of the equipment; k is the decay coefficient; for example, k = the amount of decrease in the health index in the past 30 days / 30 days.

[0112] 3) Work Order Automatic Generation Unit, which triggers the creation of an emergency work order when the health index is lower than the first threshold and the predicted maintenance time of the equipment is lower than the second threshold; or triggers the creation of a regular work order when the health index is within the preset value range;

[0113] Work order triggering conditions:

[0114] Emergency work order: Ht ≤ 0.4 or T w ≤ 7 days, automatically generate a high-priority work order.

[0115] Regular work order: 0.4 < Ht ≤ 0.6, generate a planned maintenance work order.

[0116] No need to process: Ht > 0.6, only record data and do not trigger a work order.

[0117] 4) Work Order Push Unit, which pushes the created work order to the dispatch intelligent operation center module 300.

[0118] 8. Data Prediction and Analysis Module 302, which analyzes and predicts the behavior trend of operation and maintenance personnel according to the work enthusiasm prediction model.

[0119] The operations and maintenance personnel configuration management database 200 continuously changes based on the actual situation, transmitting data to the data prediction and analysis module 302 for intelligent computing, analysis, and display. This data includes the operations and maintenance personnel's personal information, age, health status, expertise, historical order dispatch data, order retrieval data, current month's work order data, and performance. Based on this data, a model algorithm predicts and analyzes the operations and maintenance personnel's work enthusiasm, order retrieval volume, and vacation time over the next period of time.

[0120] For example, the work enthusiasm prediction is as follows:

[0121] (1) Comprehensive positivity index

[0122] At=0.4×P recen +0.3×(1-W current / W max )+0.2×H score +0.1×S match

[0123] P recen The performance average of the past three months, normalized to 0-1; W current is the current number of work orders to be processed, W max The maximum load for an individual, such as 15 times per week; H score is the health score (value ranges from 0 to 1, for example, weighted calculation based on chronic diseases, number of leave requests, etc.); S match The matching degree of work order skills is calculated by the cosine similarity between the expertise label and the work order requirements;

[0124] (2) Positiveness Attenuation Factor

[0125] λ=e -k·D (k=0.1), D represents the number of days since the last vacation.

[0126] (3) Final prediction formula

[0127] Positive prediction value J = At ​​× λ × (1 + 0.2 × I incentive )

[0128] I incentive It represents whether there is a performance reward period, and takes the value of 0 or 1; the correction value J is used as the work enthusiasm prediction score.

[0129] The mobile operation and maintenance intelligent system based on large-scale model application provided by the present invention can convert customer fault reporting voice into text, and use the large-scale model to perform semantic analysis on the text, accurately understand the customer's meaning, and recommend answers to questions for customer service personnel; use the large-scale model to accurately understand customer semantics, and automatically fill out work orders through intelligent algorithms to improve customer service work efficiency; based on the distance between the operation and maintenance engineer and the fault reporting location and the characteristics of the operation and maintenance personnel, intelligent analysis is performed and orders are assigned to the most suitable engineers, such as the top 5 engineers. These 5 engineers can independently grab orders within the scheduled 10 minutes, for example, and exert their subjective initiative; based on equipment health data, equipment failures are intelligently predicted and maintenance orders are automatically generated by the system; based on engineer data, engineer behavior is predicted and analyzed.

[0130] See also Figure 2 , introduces the intelligent mobile operation and maintenance dispatching method using the above system, including the following steps:

[0131] Step 1: The customer service representative answers the call and the voice content is transferred to the big model. The big model converts the voice content into text content and further performs semantic analysis on the text content to understand the customer's problems and needs.

[0132] Step 2: The large model recommends answers. Customer service staff answer customer questions based on the recommended answers and guide customers to solve simple questions on their own.

[0133] Step 3: If the customer can solve the problem by themselves, go to step 13 and the work order is closed; if the customer cannot solve the problem by themselves, go to step 4.

[0134] Step 4: The system intelligently fills in the work order and the customer service staff confirms it.

[0135] Step 5: Based on intelligent algorithms, the system takes into account professional attributes, work order distance, employee performance and other factors, and assigns the work order to the top 5 engineers.

[0136] Step 6: The engineer receives the work order through the mobile terminal and decides whether to grab the order based on his personal work situation.

[0137] Step 7: The system reserves 10 minutes for order grabbing. If someone grabs the order within 10 minutes, go to step 7; if no one grabs the order, go to step 5.

[0138] Step 8: The system automatically assigns the order to the engineer ranked first.

[0139] Step 9: Engineers process the work order.

[0140] Step 10: Engineers grab orders and process work orders based on their own circumstances.

[0141] Step 11: After the work order is processed, the system submits the work order.

[0142] Step 12: The user evaluates the completion status of the work order in the system.

[0143] Step 13, end.

[0144] Example: Take the "black screen" fault in a certain enterprise office area as an example:

[0145] (1) The customer calls the customer service hotline to describe the problem. The voice is converted into text by the large model and the keywords "display" and "black screen" are recognized.

[0146] (2) The system recommended the solution of "checking the connection between the power cord and the signal cable". After the customer service guided the customer to do so, the problem was not resolved.

[0147] (3) The system automatically generates a work order, and the dispatching intelligent operation center calculates the five engineers within 1 km of the fault reporting point. Among them, Engineer A (comprehensive score 92) successfully grabs the order.

[0148] (4) After arriving at the site, Engineer A confirmed that the monitor power module was faulty, replaced it, and submitted the processing record. The user gave it a 5-star rating.

[0149] (5) The system updates Engineer A’s performance data and recalculates his comprehensive score for subsequent order priority adjustments.

[0150] The present invention uses large-scale model technology to intelligently analyze customer speech semantics, recommend answers to customer service questions in real time, and automatically fill in work orders, effectively improving the work efficiency of customer service personnel. By continuously acquiring and updating operation and maintenance personnel data, equipment data, work order data, location data, etc., it can realize intelligent management of dispatching orders, which not only improves dispatching efficiency but also improves the enthusiasm of operation and maintenance personnel; by collecting historical data such as equipment design life, service life, and maintenance records, using intelligent algorithms, it predicts the time when equipment may fail, automatically initiates maintenance work orders, and improves equipment maintenance efficiency; by collecting data such as the age, expertise, and recent order grabbing of operation and maintenance personnel, it can intelligently analyze the work enthusiasm of operation and maintenance personnel and improve the management efficiency of the management level.

[0151] The various embodiments in this specification are described in a progressive manner, with each embodiment focusing on the differences from other embodiments. Reference can be made to the common and similar parts between the various embodiments. For the devices disclosed in the embodiments, since they correspond to the methods disclosed in the embodiments, the description is relatively simple, and the relevant parts can be referred to the method description.

[0152] The above description of the disclosed embodiments is intended to enable one skilled in the art to implement or use the present invention. Various modifications to these embodiments will be readily apparent to one skilled in the art, and the general principles defined herein may be implemented in other embodiments without departing from the spirit or scope of the present invention. Therefore, the present invention is not limited to the embodiments shown herein but is intended to conform to the widest scope consistent with the principles and novel features disclosed herein.

Claims

1. A mobile operation and maintenance intelligent system based on large model application, characterized by: include: Positioning database (100), used to obtain the location information of equipment and operation and maintenance personnel in real time; Operation and maintenance personnel configuration management database (200), used to store operation and maintenance personnel skills, performance and load data; Equipment configuration management database (201), used to record equipment status and maintenance history data; The large model operation center module (202) is used to convert customer speech into text and perform semantic analysis, and send customer questions and needs; The dispatch intelligent calculation center module (300) calculates the dispatch order of the top n operation and maintenance personnel with the highest scores based on the work order based on the intelligent dispatch calculation model, and activates the order grabbing mechanism; The operation and maintenance personnel mobile terminal (400) is used to obtain the operation and maintenance personnel's location information in real time and update it to the location database (100), and to support the operation and maintenance personnel's order grabbing operation and real-time update of the work order status.

2. The mobile operation and maintenance intelligent system based on large model application according to claim 1 is characterized in that: Also includes: The equipment intelligent predictive maintenance module (301) predicts equipment failures through a predictive equipment detection algorithm and generates a maintenance work order, which is then pushed to the dispatching intelligent operation center module (300).

3. The mobile operation and maintenance intelligent system based on large model application according to claim 1 is characterized in that: Also includes: The data prediction and analysis module (302) analyzes and predicts the behavior trend of operation and maintenance personnel based on the work enthusiasm prediction model.

4. The mobile operation and maintenance intelligent system based on large model application according to claim 1 is characterized in that: Also includes: The personnel information maintenance module (101) is used to add, delete, and modify the basic identity information, contact information, department and team affiliation, authority management, account credentials, shift and duty information, skills and expertise, and training record information of the operation and maintenance personnel in the operation and maintenance personnel configuration management database (200); The equipment information maintenance module (102) is used to perform addition, deletion and modification operations on the equipment usage time data, equipment maintenance time data and equipment parts replacement time data in the equipment configuration management database (201).

5. The mobile operation and maintenance intelligent system based on large model application according to claim 1 is characterized in that: Also includes: A call information processing module (103) is used to receive and process customer phone information and forward customer voice data to the large model operation center module (202); The intelligent customer service module (303) is used to receive customer questions and needs sent by the large model operation center module (202), generate recommended solutions or automatically generate work orders, and display them to customer service personnel for online answers.

6. The mobile operation and maintenance intelligent system based on large model application according to claim 1 is characterized in that: The intelligent dispatch calculation model formula in the dispatch intelligent operation center module (300) is as follows: S = α × skill matching + β × regional matching + γ × time efficiency coefficient + δ × evaluation coefficient + ∈ × order grabbing activity - ζ × current load The weight distribution is α=30%, β=25%, γ=20%, δ=15%, ∈=10%, and ζ=10%.

7. The mobile operation and maintenance intelligent system based on large model application according to claim 2 is characterized in that: The equipment intelligent predictive maintenance module (301) includes: The health index calculation unit integrates the equipment's service life, maintenance records, and sensor data to calculate the equipment's health index. The formula is as follows: Ht=w1(1-t / T)+w2(Δt / τ)+w3(s / s0) Where w1+w2+w3=1 is the weight parameter; t / T is the ratio of the equipment’s service life to its design life; Δt / τ is the ratio of the time since the last maintenance to the average maintenance interval; s / s0 is the ratio of the current sensor reading to the alarm threshold; The maintenance time prediction unit uses an exponential decay model to predict maintenance time. The formula is as follows: T w =(1-Ht) / k×T s ;T w Predict maintenance time for equipment, T s is the remaining design life of the equipment; k is the attenuation coefficient; A work order automatic generation unit triggers the creation of an emergency work order when the health index is lower than a first threshold and the equipment predicted maintenance time is lower than a second threshold; or triggers the creation of a regular work order when the health index is within a preset value range; The work order pushing unit pushes the created work order to the dispatching intelligent operation center module (300).

8. The mobile operation and maintenance intelligent system based on large model application according to claim 3 is characterized in that: The work enthusiasm prediction model formula of the data prediction analysis module (302) is as follows: Calculate the comprehensive positivity index At: At=0.4×P recen +0.3×(1-W current / W max )+0.2×H score +0.1×S match P recen The performance average of the past three months, normalized to 0-1; W current is the current number of work orders to be processed, W max is the individual's maximum load; H score Score health; S match The matching degree of work order skills is calculated by the cosine similarity between the expertise label and the work order requirements; The comprehensive positive index At combined with the attenuation factor λ = e -0.1×D Get the corrected value J: J=At×λ×(1+0.2×I incentive ) I incentive It represents whether there is a performance reward period, and takes the value of 0 or 1; the correction value J is used as the work enthusiasm prediction score.

Images