Large-model-fused airport front-end equipment operation and maintenance inspection planning method and system
By integrating large models and large language models, analyzing a variety of data and building intelligent inspection planning, the problem of failure to meet the inspection frequency and untimely failure response in the operation and maintenance management of traditional airport front-end equipment is solved, and more efficient More fair equipment inspection has improved the level of operation and maintenance management.
Patent Information
- Application Number
- CN202411456187.6
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-10-18
- Publication Date
- 2024-11-15
- Estimated Expiration
- Not applicable · inactive patent
AI Technical Summary
In the operation and maintenance management of traditional airport front-end equipment, there are problems such as the frequency of equipment inspections not meeting the standards, repeated inspection paths, large differences in personnel workloads, and untimely response to equipment failures.
Using a comprehensive large model of the airport front-end equipment operation and maintenance inspection planning method, the inspection frequency is calculated by obtaining and analyzing a variety of data, including front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, etc. , priority and path, build intelligent inspection planning, combine large language models for rule conversion and fine-tuning, and realize automatic inspection planning.
It improves inspection efficiency and fairness, reduces the probability of equipment failure, improves the refinement level of airport front-end equipment operation and maintenance management, and ensures the reasonable allocation of inspection tasks and timely maintenance of equipment.
Smart Images

Figure CN118967115A_ABST
Abstract
Description
Technical Field
[0001] The present invention belongs to the technical field of operation and maintenance management of civil aviation airport equipment, and in particular relates to a method and system for planning operation and maintenance inspection of airport front-end equipment integrating a large model. Background Art
[0002] In the traditional airport front-end equipment operation and maintenance management scenario, the front-end equipment is scattered in various areas of the airport. Different equipment has different inspection requirements due to differences in technical principles and dependencies. Airport staff rely more on their own experience for inspections, and there is no fine-grained comprehensive multi-factor automatic planning. Problems such as the inspection frequency of some equipment not meeting the standards, the inspection path being repeated back and forth, and the large differences in personnel workload often occur. At the same time, the airport operation risks are not comprehensively considered. Although a large number of manpower are arranged for inspections, sudden equipment failures and untimely responses often occur. The traditional rough operation and maintenance management model makes it difficult to perform fine-grained control according to the equipment situation. In summary, there are many pain points in the traditional airport front-end equipment operation and maintenance management scenario, which require technical improvements.
[0003] Through the above analysis, the problems and defects of the existing technology are: the inspection frequency of some equipment in the traditional airport front-end equipment operation and maintenance inspection scenario does not meet the standards, the inspection path is repeated back and forth, the workload of personnel varies greatly, and the response to equipment failure is not timely. Summary of the invention
[0004] In order to overcome the problems existing in the related art, the disclosed embodiment of the present invention provides a method and system for planning the operation and maintenance inspection of airport front-end equipment integrated with a large model. The present invention dynamically plans the operation and maintenance inspection of front-end equipment based on the operation characteristics of the airport, and uses a large model to perform more intelligent human-machine rule conversion, thereby realizing more intelligent operation and maintenance inspection intelligent planning.
[0005] The technical solution is as follows: A method for planning the operation and maintenance inspection of airport front-end equipment integrating a large model, comprising: S1: Obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; S2: The time interval range for the operation and maintenance inspection of the front-end equipment in the computer farm; S3: Calculate the future busyness of the airfield, which is characterized by the total number of flights landing at the largest apron at the same time in the future, the minimum landing interval of future flights, and the number of aircraft taxiing at the largest apron at the same time in the future; S4: Calculate the historical failure rate of the equipment to determine the frequency and priority of equipment inspections; S5: Calculate the inspection interval requirements for different equipment; S6: Calculate the working time of the inspection task; S7: Construct the inspection planning objectives of front-end equipment, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; S8: Calculate the personnel attendance and personnel skill information in the forecast interval; S9: Fine-tune the large language model of airport front-end equipment inspection rules and use the large language model for rule conversion; S10: Carry out maintenance and inspection planning for airport front-end equipment, and automatically plan according to the inspection frequency, duration and planning objectives of different equipment.
[0006] In step S1, the parsed front-end device list data is set as , set the parsed flight operation list data to , set the parsed terminal passenger distribution list data as , set the parsed historical inspection record list data as , set the parsed patrol personnel duty list data as , set the parsed inspection personnel skill list data as , set the parsed inspection plan execution data list data as , set the parsed front-end device distance matrix data to , historical inspection record list data The length of time to obtain historical data is determined by the equipment update and maintenance time to the current time; In step S1, the front-end device data is parsed, including: S101: Obtaining front-end device list data , build historical inspection record list data , traverse ; S102: Obtain specific front-end device data and the device update and maintenance time , set the current time to , then determine the time range for the device to obtain historical data is ; S103: Search the database according to the specific front-end device number and the time range determined in step S102 as conditions to obtain the historical inspection record list data, and add it to Tail, traversal completed ; In step S2, the time interval range of the front-end equipment operation and maintenance inspection planning of the computer farm is calculated, including: S201: Obtain flight operation list data , obtain the inspection plan execution data list data , construct the time interval length variable for the airport front-end equipment operation and maintenance inspection planning , construct the flight volume variable for the day , construct the variable of flight change rate executed on the day , build the pre-order planning execution manual adjustment rate variable , construct the variable of the number of flights executed on that day , construct the variable of the number of flight changes executed on that day , construct the variable of the number of tasks to be executed on the day of pre-order planning , construct the variable for manually adjusting the number of tasks for the day's pre-order planning ; S202: Traversal , get specific flight data, , determine whether the actual flight departure time is empty, if not empty, then , continue to judge the difference between the flight's planned landing time and actual landing time, and the planned take-off time and actual take-off time. When the difference between the flight's planned landing time and actual landing time or the planned take-off time and actual take-off time is ≥ 20 minutes, it is considered that the flight change range is large. , traversal completed ; S203: Traversal , obtain specific task execution data, and determine whether the actual task end time is empty. If not, then , continue to determine whether the task has been manually adjusted. If so, ; S204: Calculation and ,calculate , the specific calculation method is as follows: .
[0007] In step S3, the future operation busyness of the airfield is calculated, which is characterized by the total number of flights that will land at the largest apron at the same time in the future, the minimum landing interval of flights in the future, and the number of aircraft that will taxi at the largest apron at the same time in the future, including: S301: Obtain flight operation list data , assuming the current time is , determine the future time interval as , construct the variable of the total number of flights docked at the largest apron in the future , construct the minimum landing interval variable for future flights , construct the future maximum number of aircraft taxiing simultaneously on the apron variable , construct a variable to calculate the total number of flights that will stop at the same time in the future , construct the variable for calculating the landing interval of future flights , construct the calculation variable of the number of aircraft taxiing simultaneously on the apron in the future ; S302: Initialization and , traverse , obtain specific flight data and determine whether the actual flight check-in time is And the actual departure time of the flight is empty, if the conditions are met, then , determine the actual landing time of the flight And the actual flight check-in time is empty, if the conditions are met, then , traversal completed ; S303: Constructing an iterative time interval indicator , the initial value is 0, and the indicator is iterated +1, so the iteration time interval is , traverse , obtain specific flight data and determine whether the actual flight check-in time is If the conditions are met, then , determine whether > If the conditions are met, Numerical Assignment , determine whether the actual departure time of the flight is within If the conditions are met, then ; S304: Determine whether the actual landing time of the flight is If the conditions are met, then , determine whether If the conditions are met, Numeric assignment , determine whether the actual flight check-in time is If the conditions are met, then ; S305: , determine whether the actual landing time of the flight is If the conditions are met, the judgment > , then Assignment ,Will Set to 0; S306: Traversal completed ,judge Is it less than or equal to , if the condition is met, continue to iterate the indicator, otherwise end the iteration process; In step S4, the historical failure rate of the equipment is calculated to determine the inspection frequency and priority of the equipment, including: S401: Obtaining front-end device list data , get the historical inspection record list data , traverse ; S402: Obtain specific front-end device data, obtain the device number therein, and construct a variable for the number of front-end device failures , which is the number of equipment failures per year, and constructs the front-end equipment failure calculation variable , the initial value is 0, construct the average level variable of front-end equipment failure This variable is used to represent the average value of different fault levels of equipment in history. The equipment level is 1-4. The larger the value, the higher the fault level. Construct the front-end equipment fault level calculation variable , the initial value is 0; S403: Traversal , obtain the specific inspection record, determine whether the inspection record matches the specific front-end device, if it matches, then , + The number of fault levels in the specific inspection records, traversal completed ; S404: Calculation and , get the device update and maintenance time , assuming the current time is , the specific calculation method is as follows: S405: Numerical assignment of specific front-end device inspection priority Assign a numerical value to the historical failure rate of a specific front-end device to complete the traversal .
[0008] In step S5, the inspection interval requirements of different devices are calculated, including: S501: Obtaining front-end device list data , get terminal passenger distribution data , get the variable of the total number of flights that can stop at the maximum apron at the same time in the future , get the minimum landing interval variable for future flights , get the future maximum number of aircraft taxiing simultaneously on the apron variable , traverse ; S502: Acquire specific device data and determine the type of large area to which the device belongs; S503: Obtaining the variable of the number of front-end device failures in the specific device data , get the average level variable in the specific device data , construct the passenger density variable of the specific device location , get the latitude and longitude data in the specific device data, traverse , obtain specific passenger distribution data, determine whether the large area to which the specific device data belongs is the large area included in the obtained specific passenger distribution data, and whether the latitude and longitude data in the specific device data are within the interval range of the obtained specific passenger distribution data. If so, assign the passenger density value in the obtained specific passenger distribution data Otherwise, no processing is performed and the traversal continues. , until completion; S504: Determine the inspection time interval when the large area where the specific equipment is located is a non-flight zone area , the specific calculation method is: in, is a user-adjustable parameter. Assign inspection interval attributes in specific equipment data; S505: When the large area type to which the equipment belongs is a flight zone, determine the inspection time interval of the specific equipment , the specific calculation method is: in, is a user-adjustable parameter. Assign a value to the inspection interval attribute in the specific device data.
[0009] In step S6, the working time of the inspection task is calculated, including: S601: Obtaining front-end device list data , get the historical inspection record list data , traverse ; S602: Obtain specific front-end device data and obtain the historical average fault level therein , get the number of historical faults , get the equipment inspection type; S603: Determine the basic inspection time of the task determined by the equipment inspection type , construct the inspection task working time variable , the calculation method is as follows: S604: Calculate Assign the inspection time in the specific front-end device data to complete the traversal .
[0010] In step S7, the front-end equipment inspection planning target is constructed, including: the total number of staff members in the statistical period is , inspection task duration Indicates that the inspection driving or walking time is It means that the number of equipment inspection tasks before 6 am and after 11 pm every day is , the number of equipment inspection tasks at other times , the inspection fairness objectives are as follows: in, The weight of the user inspection task; The distance between different tasks within the statistical period is Indicates that there is You need to walk to each mission room. If you need to drive between tasks, the shortest inspection distance targets are as follows: in, The weight of the user's driving distance; The minimum number of inspection tasks is as follows: .
[0011] In step S8, the personnel attendance and personnel skill information in the prediction interval are calculated, including: S801: Obtain patrol personnel duty list data , obtain the inspection personnel skill list data , construct the forecast interval staff and staff skill structure , where Key is the employee number and Value is the employee skill tag; get the time interval length variable for front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as ; S802: Traversal , obtain the specific inspection personnel on-duty data, and determine whether the personnel are on duty or off duty. If it is within the range, continue to traverse , obtain the personnel skill data with the same personnel number, in Add a record in, the Key is the work number of the specific inspection personnel, and the Value is the personnel skill label in the obtained personnel skill data to complete the traversal .
[0012] In step S9, the large language model of the airport front-end equipment inspection rules is fine-tuned, and the large language model is used to perform rule conversion, including: first constructing a fine-tuning data set, where the fine-tuning output format is [[work number]][[(large working area, multiple areas are separated by ","), (types of equipment that can be inspected, multiple areas are separated by ","), (other skills, multiple areas are separated by ",")]], for the update of the original employee skill string, @@ indicates addition based on the original, !! indicates direct replacement and modification, and ## indicates direct deletion; The LoRA fine-tuning algorithm is used to fine-tune the large model to obtain the large language model of the airport front-end equipment inspection rules. The large model is used to process user input including: S901: Obtain text input from the user and add prompt words for large model processing; S902: Get the large model output, identify whether it contains two "[[" and "]]", if not, return to step S801 in step S8, continue to receive the large model output, if yes, get the work number in the format, add, modify, delete large areas, add, modify types, add, modify other skills; S903: Obtaining the personnel and personnel skill structure during the forecast period , use the employee number obtained in step S802 in step S8 as the key to retrieve , obtain the corresponding Value, adjust the skill string in the Value according to the addition, modification and deletion requirements output by the large model, and complete the modification.
[0013] In step S10, the airport front-end equipment operation and maintenance inspection planning is carried out, and automatic planning is carried out according to the inspection frequency, duration and planning objectives of different equipment, including: S1001: Obtain the personnel and personnel skill structure in the forecast period , get the front-end device list data , get the time interval length variable of the front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as , get the front-end device distance matrix , construct an iteration indicator , the initial value is 0, construct the subscript indicator of the staff and staff skill structure , the initial value is 0; S1002: Iteration +1, iteration time , construct the front-end equipment inspection task list for this plan The task includes the equipment number, employee number, the area to which the equipment belongs, the start time, and the end time, and constructs the staff assembly location point , used to indicate the resting place of the staff before and after the task, through the distance matrix , get the device number with the shortest distance to the staff assembly point, and search by device number , obtain specific front-end device data; S1003: Obtaining the inspection interval of the front-end device , get the inspection time of the front-end device , get the last task execution end time ,calculate Numeric ,calculate Numeric ,if ≤3 minutes, you can generate tasks for it and put the tasks into Otherwise, no processing will be done and the subsequent operation will continue; S1004: Continue through the distance matrix , get the device number with the shortest distance to the previous front-end device, and search by device number , obtain specific front-end device data, obtain , , ,calculate Numeric ,calculate Numeric , calculate the specific front-end equipment to the staff assembly location distance , calculate the distance from the specific front-end device to the previous judgment front-end device , assuming that the staff walking inspection speed , driving inspection speed , according to the distance matrix The distance traveled on the road is calculated by Numeric ,if ≥ , then generate a task for the front-end device and put the task into ; Continue processing until all front-end devices are processed; S1005: Calculation The fairness difference and total inspection distance of each staff member in the statistical period are calculated. The fairness difference is calculated by the difference between the accumulated working hours of the specific staff member and the average, and the total inspection distance is calculated by the accumulated inspection distance. First, the staff member with the smallest total inspection distance is selected. If the total inspection distances are the same, the staff member with the smallest fairness difference is selected. Get the corresponding staff member's work number and skills; S1006: Traversal , get the specific inspection task, determine whether the inspection task is empty, if it is empty, continue the subsequent operation, determine whether the staff skills meet the equipment skill requirements, equipment type, and equipment area requirements. If they meet, assign the staff member's work number to the task and calculate the distance to the staff member's previous task , the last task end time ,calculate , obtain inspection equipment data, obtain , , then the task completion time interval is , update the task end time to the last task execution end time in the inspection equipment data; S1007: Continuous traversal , until all inspection tasks have executors, continue iterating +1, judgment , then the whole process ends.
[0014] Another object of the present invention is to provide an airport front-end equipment operation and maintenance inspection planning system integrating a large model, the system implements the airport front-end equipment operation and maintenance inspection planning method integrating the large model, and the system comprises: Data acquisition and processing module: used to obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; Prediction interval calculation module: used to calculate the time interval range of the operation and maintenance inspection planning of the front-end equipment in the farm; Inspection parameter calculation module: used to calculate the future operation busyness of the airfield, represented by the total number of flights docked at the largest apron at the same time in the future, the minimum landing interval of flights in the future, and the number of aircraft taxiing at the largest apron at the same time in the future; Inspection target calculation module: used to calculate the historical failure rate of equipment and determine the inspection frequency and priority of equipment; Inspection planning data initialization module: used to calculate inspection interval requirements for different equipment; Large model processing module: used to calculate the working time of inspection tasks; Large model operation management module: used to build front-end equipment inspection planning goals, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; Inspection planning module: used to calculate the personnel working time and personnel skill information in the forecast interval; Configuration management module: used to fine-tune the large language model of airport front-end equipment inspection rules and use the large language model for rule conversion; Data output module: used for airport front-end equipment operation and maintenance inspection planning, automatic planning based on the inspection frequency, duration and planning objectives of different equipment.
[0015] In combination with all the above technical solutions, the beneficial effects of the present invention are as follows: In view of the problems existing in the traditional airport front-end equipment operation and maintenance inspection scenarios, such as the inspection frequency of some equipment not meeting the standards, the inspection paths being repeated back and forth, the large differences in personnel workload, and the untimely response to equipment failures, the present invention integrates the airport operation characteristics and historical maintenance conditions, taking into account multiple objectives such as inspection efficiency, fairness, and quality, and combines the large model to achieve more humanized and intelligent rule constraint adjustment, and finally realizes the intelligent planning of airport front-end equipment inspection, providing solid support for users to improve efficiency and reduce intensity. After application, users can refine the inspection management of airport front-end equipment, reduce the probability of inspection times not meeting the standards, achieve fairer and more efficient inspection business execution, and improve the overall management capabilities of airport front-end equipment inspection.
[0016] The present invention can comprehensively calculate the importance level of front-end equipment inspection by comprehensively considering factors such as airport operation status and historical failure rate, comprehensively considering multiple objectives such as inspection efficiency, fairness, quality, etc., and combining with large models to realize flexible configuration adjustment of rule constraints, and finally realize intelligent planning of airport front-end equipment inspection, improve the work efficiency of inspection staff, reduce part of the workload, and at the same time improve the normal operation rate of equipment. The present invention comprehensively weighs the airport operation characteristics and historical maintenance conditions, taking into account multiple objectives such as inspection efficiency, fairness, quality, etc., and combining with large models to realize more humanized and intelligent rule constraint adjustment, and finally realizes intelligent planning of airport front-end equipment inspection. BRIEF DESCRIPTION OF THE DRAWINGS
[0017] The accompanying drawings herein are incorporated in and constitute a part of the specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the principles of the present disclosure; Figure 1 It is a flow chart of the airport front-end equipment operation and maintenance inspection planning method integrating a large model provided by an embodiment of the present invention; Figure 2 It is a flow chart of calculating the time interval of historical data obtained from the historical inspection record list data provided by an embodiment of the present invention; Figure 3 It is a flow chart of calculating the time interval range for planning the operation and maintenance inspection of airport front-end equipment provided by an embodiment of the present invention; Figure 4 is a flow chart for calculating the future operation busyness of an airfield provided by an embodiment of the present invention; Figure 5 It is a flow chart of calculating the historical failure rate of equipment provided by an embodiment of the present invention; Figure 6It is a flow chart of calculating inspection interval requirements for different devices provided by an embodiment of the present invention; Figure 7 is a flow chart of calculating the working time of an inspection task provided by an embodiment of the present invention; Figure 8 It is a flow chart of rule constraint transformation using a large model provided by an embodiment of the present invention; Fig. 9 This is a flow chart of airport front-end equipment operation and maintenance inspection planning provided by an embodiment of the present invention; Fig.10 It is a schematic diagram of the airport front-end equipment operation and maintenance inspection planning system of the bright fusion large model provided by an embodiment of the present invention; Fig.11 It is a schematic diagram of the airport front-end equipment operation and maintenance inspection planning system of the bright fusion large model provided by an embodiment of the present invention; In the figure: 1. Data acquisition and processing module; 2. Prediction interval calculation module; 3. Inspection parameter calculation module; 4. Inspection target calculation module; 5. Inspection planning data initialization module; 6. Large model processing module; 7. Large model operation management module; 8. Inspection planning module; 9. Configuration management module; 10. Data output module. DETAILED DESCRIPTION
[0018] In order to make the above-mentioned objects, features and advantages of the present invention more obvious and easy to understand, the specific embodiments of the present invention are described in detail below in conjunction with the accompanying drawings. In the following description, many specific details are set forth to facilitate a full understanding of the present invention. However, the present invention can be implemented in many other ways different from those described herein, and those skilled in the art can make similar improvements without violating the connotation of the present invention, so the present invention is not limited by the specific implementation disclosed below.
[0019] The innovation of the present invention lies in: in order to address the problems existing in the traditional airport front-end equipment operation and maintenance inspection scenarios, such as the inspection frequency of some equipment not meeting the standards, the inspection paths being repeated, the large differences in personnel workload, and the untimely response to equipment failures, the present invention integrates data from multiple parties to establish an inspection interval calculation method that is consistent with the current status of airport production and operation and the historical failure rate, comprehensively considers factors such as inspection efficiency, fairness, and quality, and uses a large model to make more intelligent and humane constraint factor adjustments, ultimately achieving intelligent planning for airport front-end equipment inspections, improving the level of refined management of airport front-end equipment operation and maintenance, and providing solid support for users to improve efficiency and fairness and reduce intensity.
[0020] Embodiment 1, as Figure 1 As shown in the figure, the airport front-end equipment operation and maintenance inspection planning method integrating the big model includes: S1: Obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; The front-end equipment data includes equipment number, equipment name, equipment longitude and latitude, equipment floor, equipment region, equipment skill requirements, equipment type, inspection interval, inspection priority, number of historical faults, inspection duration, last task execution end time, equipment update and maintenance time. The flight operation data includes flight number, flight number, planned landing time, planned take-off time, actual landing time, actual take-off time, actual check-in time, actual check-out time, flight status, flight parking space, and flight boarding gate. The passenger distribution data of the terminal includes the passenger region, passenger area range, and passenger density. The historical inspection record data includes record number, record time, inspection equipment, fault level, and handling opinions. The inspection personnel duty data includes personnel number, personnel working time, and personnel off-duty time. The inspection personnel skill data includes personnel number, personnel employee number, personnel name, and personnel skill label. The inspection plan execution data includes task number, equipment number, employee number, task start time, and task end time. The front-end equipment distance matrix data includes number, source number, terminal number, distance, and driving mode. Set the parsed front-end device list data as , set the parsed flight operation list data to , set the parsed terminal passenger distribution list data as , set the parsed historical inspection record list data as , set the parsed patrol personnel duty list data as , set the parsed inspection personnel skill list data as , set the parsed inspection plan execution data list data as , set the parsed front-end device distance matrix data to , historical inspection record list data The length of time for obtaining historical data is determined by the equipment update and maintenance time to the current time. Because after the update and maintenance, the equipment operation status will be greatly different from before, obtaining the previous historical data to calculate the average monthly failure number will have a large inaccuracy, such as Figure 2 , The specific calculation method is as follows: S101: Obtaining front-end device list data , build historical inspection record list data , traverse ; S102: Obtain specific front-end device data and the device update and maintenance time , set the current time to , then determine the time range for the device to obtain historical data is ; S103: Search the database according to the specific front-end device number and the time range determined in step S102 as conditions to obtain the historical inspection record list data, and add it to Tail, traversal completed ; S2: The time interval range for the operation and maintenance inspection of the front-end equipment in the computer farm; This time range is determined by the number of flights, flight changes, and the execution of previous plans, such as Figure 3 , the specific calculation method is as follows: S201: Obtain flight operation list data , obtain the inspection plan execution data list data , construct the time interval length variable for the airport front-end equipment operation and maintenance inspection planning , construct the flight volume variable for the day , construct the variable of flight change rate executed on the day , build the pre-order planning execution manual adjustment rate variable , construct the variable of the number of flights executed on that day , construct the variable of the number of flight changes executed on that day , construct the variable of the number of tasks to be executed on the day of pre-order planning , construct the variable for manually adjusting the number of tasks for the day's pre-order planning ; S202: Traversal , get specific flight data, , determine whether the actual flight departure time is empty, if not empty, then , continue to judge the difference between the flight's planned landing time and actual landing time, and the planned take-off time and actual take-off time. When the difference between the flight's planned landing time and actual landing time or the planned take-off time and actual take-off time is ≥ 20 minutes, it is considered that the flight change range is large. , traversal completed ; S203: Traversal , obtain specific task execution data, and determine whether the actual task end time is empty. If not, then , continue to determine whether the task has been manually adjusted. If so, ; S204: Calculation and ,calculate , the specific calculation method is as follows: S3: Calculate the future busyness of the airfield, which is characterized by the total number of flights landing at the largest apron at the same time in the future, the minimum landing interval of future flights, and the number of aircraft taxiing at the largest apron at the same time in the future; like Figure 4 , the specific calculation method is as follows: S301: Obtain flight operation list data , assuming the current time is , determine the future time interval as , construct the variable of the total number of flights docked at the largest apron in the future , construct the minimum landing interval variable for future flights , construct the future maximum number of aircraft taxiing simultaneously on the apron variable , construct a variable to calculate the total number of flights that will stop at the same time in the future , construct the variable for calculating the landing interval of future flights , construct the calculation variable of the number of aircraft taxiing simultaneously on the apron in the future ; S302: Initialization and , traverse , obtain specific flight data and determine whether the actual flight check-in time is And the actual departure time of the flight is empty, if the conditions are met, then , determine the actual landing time of the flight And the actual flight check-in time is empty, if the conditions are met, then , traversal completed ; S303: Constructing an iteration time interval indicator , the initial value is 0, and the indicator is iterated +1, so the iteration time interval is , traverse , obtain specific flight data and determine whether the actual flight check-in time is If the conditions are met, then , determine whether > If the conditions are met, Numerical Assignment , determine whether the actual departure time of the flight is within If the conditions are met, then ; S304: Determine whether the actual landing time of the flight is If the conditions are met, then , determine whether If the conditions are met, Numeric assignment , determine whether the actual flight check-in time is If the conditions are met, then ; S305: , determine whether the actual landing time of the flight is If the conditions are met, the judgment > , then Assignment ,Will Set to 0; S306: Traversal completed ,judge Is it less than or equal to , if the condition is met, continue to iterate the indicator, otherwise end the iteration process; S4: Calculate the historical failure rate of the equipment to determine the frequency and priority of equipment inspections; This is used to determine the frequency and priority of equipment inspections in the future. Figure 5 , the specific calculation method is as follows: S401: Obtaining front-end device list data , get the historical inspection record list data , traverse ; S402: Obtain specific front-end device data, obtain the device number therein, and construct a variable for the number of front-end device failures , which is the number of equipment failures per year, and constructs the front-end equipment failure calculation variable , the initial value is 0, construct the average level variable of front-end equipment failure This variable is used to represent the average value of different fault levels of equipment in history. The equipment level is 1-4. The larger the value, the higher the fault level. Construct the front-end equipment fault level calculation variable , the initial value is 0; S403: Traversal , obtain the specific inspection record, determine whether the inspection record matches the specific front-end device, if it matches, then , + The number of fault levels in the specific inspection records, traversal completed ; S404: Calculation and , get the device update and maintenance time , assuming the current time is , the specific calculation method is as follows: S405: Numerical assignment of specific front-end device inspection priority Assign a numerical value to the historical failure rate of a specific front-end device to complete the traversal .
[0021] S5: Calculate the inspection interval requirements for different equipment; The inspection interval is related to the large area to which the equipment belongs, the passenger density, the future busyness of the flight area, and the historical failure rate of the equipment, such as Figure 6 , the specific calculation method is: S501: Obtaining front-end device list data , get terminal passenger distribution data , get the variable of the total number of flights that can stop at the maximum apron at the same time in the future , get the minimum landing interval variable for future flights , get the future maximum number of aircraft taxiing simultaneously on the apron variable , traverse ; S502: Acquire specific device data and determine the type of large area to which the device belongs; Determine the type of large area to which the device belongs. The specific area types are divided as shown in Table 1.
[0022] Table 1 Regional classification When the large area type to which the equipment belongs is not an airfield, the passenger density distribution needs to be considered when calculating the equipment inspection interval requirements.
[0023] S503: Obtaining the variable of the number of front-end device failures in the specific device data , get the average level variable in the specific device data , construct the passenger density variable of the specific device location (Using the average distance between passengers, in meters), obtain the latitude and longitude data in the specific device data, and traverse , obtain specific passenger distribution data, determine whether the large area to which the specific device data belongs is the large area included in the obtained specific passenger distribution data, and whether the latitude and longitude data in the specific device data are within the interval range of the obtained specific passenger distribution data. If so, assign the passenger density value in the obtained specific passenger distribution data Otherwise, no processing is performed and the traversal continues. , until completed.
[0024] S504: Determine the inspection time interval when the large area where the specific equipment is located is a non-flight zone area , the specific calculation method is: in, is a user-adjustable parameter. Assign inspection interval attributes in specific equipment data; S505: When the large area type to which the equipment belongs is a flight zone, determine the inspection time interval of the specific equipment , the specific calculation method is: in, is a user-adjustable parameter. Assign a value to the inspection interval attribute in the specific device data.
[0025] S6: Calculate the working time of the inspection task; The working time is related to the historical average fault level of the equipment, the number of historical faults, and the type of equipment inspection, which provides a basis for subsequent inspection planning, such as Figure 7 , the specific calculation process is as follows: S601: Obtaining front-end device list data , get the historical inspection record list data , traverse .
[0026] S602: Obtain specific front-end device data and obtain the historical average fault level therein , get the number of historical faults , obtain the equipment inspection type, the inspection equipment type is shown in Table 2; Table 2 Inspection equipment types S603: Determine the basic inspection time of the task determined by the equipment inspection type , construct the inspection task working time variable , the calculation method is as follows: S604: Calculate Assign the inspection time in the specific front-end device data. Complete the traversal .
[0027] S7: Construct the inspection planning objectives of front-end equipment, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; Compared with the traditional fairness characterized by the balance of the number of devices inspected, it is more reasonable to characterize inspection fairness by the balance of employee working hours, because many inspection tasks require driving or walking for a long distance, so the employee working hours include the driving or walking time from one task to another. At the same time, in order to reflect fairness, the equipment inspection staff work harder in the early morning and at night, and more proportional weights are added to the equipment inspection working hours before 6 am and after 11 pm every day. Assuming that the total number of staff in the statistical period is , inspection task duration Indicates that the inspection driving or walking time is It means that the number of equipment inspection tasks before 6 am and after 11 pm every day is , the number of equipment inspection tasks at other times , the inspection fairness objectives are as follows: in, It is the inspection task weight of the user; the user can adjust the inspection task weight before 6 am and after 11 pm every day.
[0028] Assume that the distance between different tasks in the statistical period is It means that because the inspection equipment is spread across different areas of the airport, some inspections need to be carried out by car, while some inspections need to be carried out on foot. The distances represented by the two will have different weights in the calculation of the target. Assume that You need to walk to each mission room. If you need to drive between tasks, the shortest inspection distance targets are as follows: in, The weight of the user's driving distance; The minimum number of inspection tasks is as follows: S8: Calculate the personnel attendance and personnel skill information in the forecast interval; This step provides the basis for the subsequent front-end equipment inspection planning. The specific steps are as follows: S801: Obtain patrol personnel duty list data , obtain the inspection personnel skill list data , construct the forecast interval staff and staff skill structure , where the Key is the employee number and the Value is the employee skill tag. Get the time interval length variable for the front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as .
[0029] S802: Traversal , obtain the specific inspection personnel on-duty data, and determine whether the personnel are on duty or off duty. If it is within the range, continue to traverse , obtain the personnel skill data with the same personnel number, in Add a record in, the Key is the work number of the specific inspection personnel, and the Value is the personnel skill label in the obtained personnel skill data. Complete the traversal .
[0030] S9: Fine-tune the large language model of airport front-end equipment inspection rules and use the large language model for rule conversion; First, construct a fine-tuning dataset. The fine-tuning samples are shown in Table 3. Table 3 Fine-tuning examples The fine-tuning output format is [[employee number]][[(large working area, multiple characters separated by “,”),(type of equipment that can be inspected, multiple characters separated by “,”),(other skills, multiple characters separated by “,”)]]; for the update of the original employee skill string, “@@” indicates addition based on the original, “!!” indicates direct replacement and modification, and “##” indicates direct deletion.
[0031] The LoRA fine-tuning algorithm is used to fine-tune the large model to obtain the large language model of the airport front-end equipment inspection rules. The large model is used to process user input, such as Figure 8 , the specific steps are as follows: S901: Obtain the text input of the user and add prompt words for large model processing. The specific prompt words are shown in Table 4; Table 4 Prompt words S902: Get the large model output and identify whether it contains 2 "[[" and "]]". If not, go to S801 and continue to receive the large model output. If yes, get the work number in the format, add, modify or delete large areas, add modification types, and add or modify other skills.
[0032] S903: Obtaining the personnel and personnel skill structure during the forecast period , use the employee number obtained in S802 as the key to retrieve , obtain the corresponding Value, adjust the skill string in the Value according to the addition, modification and deletion requirements output by the large model, and complete the modification.
[0033] S10: Carry out the operation and maintenance inspection planning of the front-end equipment of the airport, and automatically plan according to the inspection frequency, duration and planning objectives of different equipment, such as Fig. 9 , the specific steps are as follows: S1001: Obtain the personnel and personnel skill structure in the forecast period , get the front-end device list data , get the time interval length variable of the front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as , get the front-end device distance matrix , construct an iteration indicator , the initial value is 0, construct the subscript indicator of the staff and staff skill structure , the initial value is 0.
[0034] S1002: Iteration +1, iteration time , construct the front-end equipment inspection task list for this plan The task includes the equipment number, employee number, the area to which the equipment belongs, the start time, and the end time, and constructs the staff assembly location point , used to indicate the resting place of the staff before and after the task, through the distance matrix , get the device number with the shortest distance to the staff assembly point, and search by device number , get the specific front-end device data.
[0035] S1003: Obtaining the inspection interval of the front-end device , get the inspection time of the front-end device , get the last task execution end time ,calculate Numeric ,calculate Numeric ,if ≤3 minutes, you can generate tasks for it and put the tasks into Otherwise, no processing will be done and the subsequent operation will continue.
[0036] S1004: Continue through the distance matrix , get the device number with the shortest distance to the previous front-end device, and search by device number , obtain specific front-end device data, obtain , , ,calculate Numeric ,calculate Numeric , calculate the specific front-end equipment to the staff assembly location distance , calculate the distance from the specific front-end device to the previous judgment front-end device , assuming that the staff walking inspection speed , driving inspection speed , according to the distance matrix The distance traveled on the road is calculated by Numeric ,if ≥ , then generate a task for the front-end device and put the task into Continue processing until all front-end devices have been processed.
[0037] S1005: Calculation The fairness difference and total inspection distance of each staff member in the statistical period are calculated. The fairness difference is calculated by the difference between the accumulated working hours of the specific staff member and the average, and the total inspection distance is calculated by the accumulated inspection distance. First, the staff member with the smallest total inspection distance is selected. If the total inspection distances are the same, the staff member with the smallest fairness difference is selected. This calculation method is to meet the inspection fairness and the shortest inspection distance goals as much as possible. Get the corresponding staff member’s work number and skills.
[0038] S1006: Traversal , get the specific inspection task, determine whether the inspection task is empty, if it is empty, continue the subsequent operation, determine whether the staff skills meet the equipment skill requirements, equipment type, and equipment area requirements. If they meet, assign the staff member's work number to the task and calculate the distance to the staff member's previous task , the last task end time ,calculate , obtain inspection equipment data, obtain , , then the task completion time interval is , update the task end time to the last task execution end time in the inspection equipment data.
[0039] S1007: Continuous traversal , until all inspection tasks have executors, continue iterating +1, judgment > , then the whole process ends.
[0040] Embodiment 2, as Fig.10As shown in the figure, the airport front-end equipment operation and maintenance inspection planning system integrating the large model includes: Data acquisition and processing module 1, used to obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; Prediction interval calculation module 2 is used to calculate the time interval range of the operation and maintenance inspection planning of the front-end equipment of the farm; Inspection parameter calculation module 3 is used to calculate the future operation busyness of the airfield, which is characterized by the total number of flights landing at the largest apron at the same time in the future, the minimum landing interval of flights in the future, and the number of aircraft taxiing at the largest apron at the same time in the future; Inspection target calculation module 4, used to calculate the historical failure rate of equipment, and to determine the inspection frequency and priority of equipment; Inspection planning data initialization module 5, used to calculate inspection interval requirements for different equipment; Large model processing module 6, used to calculate the working time of the inspection task; The large model operation management module 7 is used to construct the front-end equipment inspection planning objectives, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; Inspection planning module 8 is used to calculate the personnel attendance and personnel skill information in the forecast interval; Configuration management module 9, used to fine-tune the large language model of airport front-end equipment inspection rules and use the large language model to perform rule conversion; The data output module 10 is used to carry out the operation and maintenance inspection planning of the airport front-end equipment, and automatically plan according to the inspection frequency, duration and planning objectives of different equipment.
[0041] in, Fig.11 This is the schematic diagram of the airport front-end equipment operation and maintenance inspection planning system that integrates the large model.
[0042] It can be seen from the above embodiments that the present invention proposes a multi-factor airport front-end equipment operation and maintenance inspection planning method and system that integrates a large model. In view of the problems existing in the traditional airport front-end equipment operation and maintenance inspection scenarios, such as the inspection frequency of some equipment not meeting the standards, the inspection path being repeated back and forth, the large difference in personnel workload, and the untimely response to equipment failures, the present invention comprehensively weighs the airport operation characteristics and historical maintenance conditions, taking into account multiple objectives such as inspection efficiency, fairness, and quality, and at the same time combines the large model to achieve more humane and intelligent rule constraint adjustment, ultimately realizing intelligent planning of airport front-end equipment inspections, and providing solid support for users to improve efficiency and reduce intensity.
[0043] In order to effectively demonstrate the positive effects that the present invention can achieve, it is compared with the traditional airport front-end equipment inspection and operation and maintenance scenario. In the traditional scenario, the business department formulates a relatively rough inspection plan based on the front-end facility operation status and geographical location characteristics, such as how long to inspect once, and the inspection time interval of the front-end equipment on an inspection route is basically the same, and the staff conducts inspections as required. The present invention will first determine the inspection intervals of different front-end equipment according to the airport's on-site operation status and historical failure rate data to ensure that the probability of front-end equipment failure and the impact of failure on production can be minimized. Then, the present invention will comprehensively consider efficiency, fairness, quality and other multi-objective factors under the conditions of different equipment inspection time interval requirements and skill constraints, and give a fine-grained inspection task scheduling scheme. Compared with the traditional scheme, the present invention scheme has a more fine-grained control of the front-end inspection equipment, and at the same time gives a refined intelligent inspection scheduling time. In comparison, it has more application advantages, can improve the stable operation level of airport front-end equipment, and improve the refined management level of airport inspection and operation and maintenance.
[0044] The above description is only a preferred specific implementation manner of the present invention, but the protection scope of the present invention is not limited thereto. Any modifications, equivalent substitutions and improvements made by any technician familiar with the technical field within the technical scope disclosed by the present invention and within the spirit and principles of the present invention should be covered within the protection scope of the present invention.
Claims
1. A method for planning the operation and maintenance inspection of airport front-end equipment integrating a large model, characterized in that: The method includes: S1: Obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; S2: The time interval range for the operation and maintenance inspection of the front-end equipment in the computer farm; S3: Calculate the future busyness of the airfield, which is characterized by the total number of flights landing at the largest apron at the same time in the future, the minimum landing interval of future flights, and the number of aircraft taxiing at the largest apron at the same time in the future; S4: Calculate the historical failure rate of the equipment to determine the frequency and priority of equipment inspections; S5: Calculate the inspection interval requirements for different equipment; S6: Calculate the working time of the inspection task; S7: Construct the inspection planning objectives of front-end equipment, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; S8: Calculate the personnel attendance and personnel skill information in the forecast interval; S9: Fine-tune the large language model of airport front-end equipment inspection rules and use the large language model for rule conversion; S10: Carry out maintenance and inspection planning for airport front-end equipment, and automatically plan according to the inspection frequency, duration and planning objectives of different equipment.
2. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S1, the parsed front-end device list data is set as , set the parsed flight operation list data to , set the parsed terminal passenger distribution list data as , set the parsed historical inspection record list data as , set the parsed patrol personnel duty list data as , set the parsed inspection personnel skill list data as , set the parsed inspection plan execution data list data as , set the parsed front-end device distance matrix data to , historical inspection record list data The length of time to obtain historical data is determined by the equipment update and maintenance time to the current time; In step S1, the front-end device data is parsed, including: S101: Obtaining front-end device list data , build historical inspection record list data , traverse ; S102: Obtain specific front-end device data and the device update and maintenance time , set the current time to , then determine the time range for the device to obtain historical data is ; S103: Search the database according to the specific front-end device number and the time range determined in step S102 as conditions to obtain the historical inspection record list data, and add it to Tail, traversal completed ; In step S2, the time interval range of the front-end equipment operation and maintenance inspection planning of the computer farm is calculated, including: S201: Obtain flight operation list data , obtain the inspection plan execution data list data , construct the time interval length variable for the airport front-end equipment operation and maintenance inspection planning , construct the flight volume variable for the day , construct the variable of flight change rate executed on the day , build the pre-order planning execution manual adjustment rate variable , construct the variable of the number of flights executed on that day , construct the variable of the number of flight changes executed on that day , construct the variable of the number of tasks to be executed on the day of pre-order planning , construct the variable for manually adjusting the number of tasks for the day's pre-order planning ; S202: Traversal , get specific flight data, , determine whether the actual flight departure time is empty, if not empty, then , continue to judge the difference between the flight's planned landing time and actual landing time, and the planned take-off time and actual take-off time. When the difference between the flight's planned landing time and actual landing time or the planned take-off time and actual take-off time is ≥ 20 minutes, it is considered that the flight change range is large. , traversal completed ; S203: Traversal , obtain specific task execution data, and determine whether the actual task end time is empty. If not, then , continue to determine whether the task has been manually adjusted. If so, ; S204: Calculation and ,calculate , the specific calculation method is as follows: 。 3. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S3, the future operation busyness of the airfield is calculated, which is characterized by the total number of flights that will dock at the largest apron at the same time in the future, the minimum landing interval of flights in the future, and the number of aircraft that will taxi at the largest apron at the same time in the future, including: S301: Obtain flight operation list data , assuming the current time is , determine the future time interval as , construct the variable of the total number of flights docked at the largest apron in the future , construct the minimum landing interval variable for future flights , construct the future maximum number of aircraft taxiing simultaneously on the apron variable , construct a variable to calculate the total number of flights that will stop at the same time in the future , construct the variable for calculating the landing interval of future flights , construct the calculation variable of the number of aircraft taxiing simultaneously on the apron in the future ; S302: Initialization and , traverse , obtain specific flight data and determine whether the actual flight check-in time is And the actual departure time of the flight is empty, if the conditions are met, then , determine the actual landing time of the flight And the actual flight check-in time is empty, if the conditions are met, then , traversal completed ; S303: Constructing an iterative time interval indicator , the initial value is 0, and the indicator is iterated +1, so the iteration time interval is , traverse , obtain specific flight data and determine whether the actual flight check-in time is If the conditions are met, then , determine whether > If the conditions are met, Numeric assignment , to determine whether the actual departure time of the flight is within If the conditions are met, then ; S304: Determine whether the actual landing time of the flight is If the conditions are met, then , determine whether If the conditions are met, Numeric assignment , determine whether the actual flight check-in time is If the conditions are met, then ; S305: , determine whether the actual landing time of the flight is If the conditions are met, the judgment > , then Assignment ,Will Set to 0; S306: Traversal completed ,judge Is it less than or equal to , if the condition is met, continue to iterate the indicator, otherwise end the iteration process; In step S4, the historical failure rate of the equipment is calculated to determine the inspection frequency and priority of the equipment, including: S401: Obtaining front-end device list data , get the historical inspection record list data , traverse ; S402: Obtain specific front-end device data, obtain the device number therein, and construct a variable for the number of front-end device failures , which is the number of equipment failures per year, and constructs the front-end equipment failure calculation variable , the initial value is 0, construct the average level variable of front-end equipment failure This variable is used to represent the average value of different fault levels of equipment in history. The equipment level is 1-4. The larger the value, the higher the fault level. Construct the front-end equipment fault level calculation variable , the initial value is 0; S403: Traversal , obtain the specific inspection record, determine whether the inspection record matches the specific front-end device, if it matches, then , + The number of fault levels in the specific inspection records, traversal completed ; S404: Calculation and , get the device update and maintenance time , assuming the current time is , the specific calculation method is as follows: S405: Numerical assignment of specific front-end device inspection priority Assign a numerical value to the historical failure rate of a specific front-end device to complete the traversal .
4. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S5, the inspection interval requirements of different devices are calculated, including: S501: Obtaining front-end device list data , get terminal passenger distribution data , get the variable of the total number of flights that can stop at the maximum apron at the same time in the future , get the minimum landing interval variable for future flights , get the future maximum number of aircraft taxiing simultaneously on the apron variable , traverse ; S502: Acquire specific device data and determine the type of large area to which the device belongs; S503: Obtaining the variable of the number of front-end device failures in the specific device data , get the average level variable in the specific device data , construct the passenger density variable of the specific device location , get the latitude and longitude data in the specific device data, traverse , obtain specific passenger distribution data, determine whether the large area to which the specific device data belongs is the large area included in the obtained specific passenger distribution data, and whether the latitude and longitude data in the specific device data are within the interval range of the obtained specific passenger distribution data. If so, assign the passenger density value in the obtained specific passenger distribution data Otherwise, no processing is performed and the traversal continues. , until completion; S504: Determine the inspection time interval when the large area where the specific equipment is located is a non-flight zone area , the specific calculation method is: in, is a user-adjustable parameter. Assign inspection interval attributes in specific equipment data; S505: When the large area type to which the equipment belongs is a flight zone, determine the inspection time interval of the specific equipment , the specific calculation method is: in, is a user-adjustable parameter. Assign a value to the inspection interval attribute in the specific device data.
5. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S6, the working time of the inspection task is calculated, including: S601: Obtaining front-end device list data , get the historical inspection record list data , traverse ; S602: Obtain specific front-end device data and obtain the historical average fault level therein , get the number of historical faults , get the equipment inspection type; S603: Determine the basic inspection time of the task determined by the equipment inspection type , construct the inspection task working time variable , the calculation method is as follows: S604: Calculate Assign the inspection time in the specific front-end device data to complete the traversal .
6. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S7, the front-end equipment inspection planning target is constructed, including: the total number of staff members in the statistical period is , inspection task duration Indicates that the inspection driving or walking time is It means that the number of equipment inspection tasks before 6 am and after 11 pm every day is , the number of equipment inspection tasks at other times , the inspection fairness objectives are as follows: in, The weight of the user inspection task; The distance between different tasks within the statistical period is Indicates that there is You need to walk to each mission room. If you need to drive between tasks, the shortest inspection distance targets are as follows: in, The weight of the user's driving distance; The minimum number of inspection tasks is as follows: 。 7. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S8, the personnel attendance and personnel skill information in the prediction interval are calculated, including: S801: Obtain patrol personnel duty list data , obtain the inspection personnel skill list data , construct the forecast interval staff and staff skill structure , where Key is the employee number and Value is the employee skill tag; get the time interval length variable for front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as ; S802: Traversal , obtain the specific inspection personnel on-duty data, and determine whether the personnel are on duty or off duty. If it is within the range, continue to traverse , obtain the personnel skill data with the same personnel number, in Add a record in, the Key is the work number of the specific inspection personnel, and the Value is the personnel skill label in the obtained personnel skill data to complete the traversal .
8. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S9, the large language model of the airport front-end equipment inspection rules is fine-tuned, and the large language model is used for rule conversion, including: fine-tuning the large model using the LoRA fine-tuning algorithm to obtain the large language model of the airport front-end equipment inspection rules, and using the large model to process user input, specifically including the following steps: S901: Obtain text input from the user and add prompt words for large model processing; S902: Get the large model output, identify whether it contains "[[" and "]]" are both 2, if not, return to step S801 in step S8, continue to receive the large model output, if yes, get the work number in the format, add, modify, delete the large area, add, modify the type, add, modify other skills; S903: Obtaining the personnel and personnel skill structure during the forecast period , use the employee number obtained in step S802 as the key to search , obtain the corresponding Value, adjust the skill string in the Value according to the addition, modification and deletion requirements output by the large model, and complete the modification.
9. The airport front-end equipment operation and maintenance inspection planning method integrating a large model according to claim 1 is characterized in that: In step S10, the airport front-end equipment operation and maintenance inspection planning is carried out, and automatic planning is carried out according to the inspection frequency, duration and planning objectives of different equipment, including: S1001: Obtain the personnel and personnel skill structure in the forecast period , get the front-end device list data , get the time interval length variable of the front-end equipment operation and maintenance inspection planning , assuming the current time is , determine the future time interval as , get the front-end device distance matrix , construct an iteration indicator , the initial value is 0, construct the subscript indicator of the staff and staff skill structure , the initial value is 0; S1002: Iteration +1, iteration time , construct the front-end equipment inspection task list for this plan The task includes the equipment number, employee number, the area to which the equipment belongs, the start time, and the end time, and constructs the staff assembly location point , used to indicate the resting place of the staff before and after the task, through the distance matrix , get the device number with the shortest distance to the staff assembly point, and search by device number , obtain specific front-end device data; S1003: Obtaining the inspection interval of the front-end device , get the inspection time of the front-end device , get the last task execution end time ,calculate Numeric ,calculate Numeric ,if ≤3 minutes, generate a task and put it into Otherwise, no processing will be done and the subsequent operation will continue; S1004: Continue through the distance matrix , get the device number with the shortest distance to the previous front-end device, and search by device number , obtain specific front-end device data, obtain , , ,calculate Numeric ,calculate Numeric , calculate the specific front-end equipment to the staff assembly location distance , calculate the distance from the specific front-end device to the previous judgment front-end device , assuming that the staff's walking inspection speed is , driving inspection speed , according to the distance matrix The distance traveled on the road is calculated by Numeric ,if ≥ , then generate a task for the front-end device and put the task into ; Continue processing until all front-end devices are processed; S1005: Calculation The fairness difference and total inspection distance of each staff member in the statistical period are calculated. The fairness difference is calculated by the difference between the accumulated working hours of the specific staff member and the average, and the total inspection distance is calculated by the accumulated inspection distance. First, the staff member with the smallest total inspection distance is selected. If the total inspection distances are the same, the staff member with the smallest fairness difference is selected. Get the corresponding staff member's work number and skills; S1006: Traversal , get the specific inspection task, determine whether the inspection task is empty, if it is empty, continue the subsequent operation, determine whether the staff skills meet the equipment skill requirements, equipment type, and equipment area requirements. If they meet, assign the staff member's work number to the task and calculate the distance to the staff member's previous task , the last task end time ,calculate , obtain inspection equipment data, obtain , , then the task completion time interval is , update the task end time to the last task execution end time in the inspection equipment data; S1007: Continuous traversal , until all inspection tasks have executors, continue iterating +1, judgment , then the whole process ends.
10. An airport front-end equipment operation and maintenance inspection planning system integrating a large model, characterized in that: The system implements the airport front-end equipment operation and maintenance inspection planning method integrating the large model as described in any one of claims 1 to 9, and the system includes: Data acquisition and processing module (1): used to obtain and analyze front-end equipment data, flight operation data, terminal passenger distribution data, historical inspection record data, inspection personnel duty data, inspection personnel skill data, inspection plan execution data, and front-end equipment distance matrix data; Prediction interval calculation module (2): used to calculate the time interval range of the operation and maintenance inspection planning of the front-end equipment in the farm; Inspection parameter calculation module (3): used to calculate the future operation busyness of the airfield, represented by the total number of flights landing at the largest apron at the same time in the future, the minimum landing interval of flights in the future, and the number of aircraft taxiing at the largest apron at the same time in the future; Inspection target calculation module (4): used to calculate the historical failure rate of equipment and to determine the inspection frequency and priority of equipment; Inspection planning data initialization module (5): used to calculate the inspection interval requirements for different equipment; Large model processing module (6): used to calculate the working time of the inspection task; Large model operation management module (7): used to construct front-end equipment inspection planning objectives, including inspection fairness, shortest inspection distance, and minimum number of inspection tasks; Inspection planning module (8): used to calculate the personnel attendance and personnel skill information in the forecast interval; Configuration management module (9): used to fine-tune the large language model of airport front-end equipment inspection rules and use the large language model to perform rule conversion; Data output module (10): used to carry out airport front-end equipment operation and maintenance inspection planning, and automatically plan according to the inspection frequency, duration and planning objectives of different equipment.
Citation Information
Patent Citations
Civil aviation airport ground service support personnel scheduling method and system thereof
CN109165782A
Intelligent scheduling method based on machine learning and operation planning optimization
CN116934046A
Power plant equipment intelligent inspection method and system based on cloud side-end cooperation
CN117014472A
Continuous unmanned aerial vehicle airport inspection scheduling method
CN117369516A
Unmanned aerial vehicle inspection management system based on gridding division
CN117521932A
Cited By
Airport flying area water surface floating photovoltaic power generation system test method and platform
CN119727598A
