A flight crew scheduling method for large aircraft public air transport

Abstract

The present invention provides a flight crew scheduling method for large aircraft public air transport, including constructing a large aircraft flight crew scheduling model; and obtaining a flight crew scheduling method by solving the model. The model constructed by the present invention includes an aircraft allocation model and a flight crew scheduling model. The aircraft allocation model can improve the utilization rate of aircraft by setting long-distance route priorities, that is, the same aircraft can be assigned more flight missions than actual ones and achieve a longer total flight time. The longer total flight time will also reduce the cost index CI, thereby reducing the total cost of the airline. In the flight crew scheduling model, by setting bases with a larger proportion of pilots as priorities, the complex situation of changing aircraft models due to insufficient number of pilots can be effectively reduced, and at the same time, the problem of excessively high cost of dispatching pilots due to insufficient number of pilots can be reduced.

Description

Technical Field

[0001] The invention belongs to the field of aerospace technology, and in particular relates to a flight crew scheduling method in public air transportation of large aircraft. Background Art

[0002] Traditional crew scheduling is mainly based on the determined flight segments to compile a set of crew schedules that depart from a certain base and can return to the base within a fixed period. The scheduling needs to meet the regulations on flight time, duty time and rest time of civil aviation transport flight personnel issued by the Civil Aviation Administration. The scheduled crew must cover all flight segments of the flight plan so that the scheduling plan meets the constraints of legality and feasibility. For example, Zhang Guangwei constructed a crew task pairing mathematical model based on the improved dynamic cultural gene algorithm with the goal of generating a set of crew task rings with the lowest cost, covering all flights and meeting the statutory standards, and tested it through experimental data. Through the comparison of multiple key performance indicators, a higher-quality crew scheduling plan was obtained. Xiang Dubing constructed a task-crew matching model and solved it with a mixed sub-problem of local search and precise search. The results show that the research method can solve the complex combinatorial optimization problem of large-scale crew scheduling in a short time. Zhang Mi constructed a single-base crew scheduling model based on mathematical programming, added constraints on dual airports and flights, and designed a heuristic algorithm. The results show that the structure and optimization effect of the model are good. Lan Boxiong proposed a mathematical programming model for crew scheduling that takes into account random delay factors with the goal of minimizing the cost of each task and the cost of delay, and proposed a heuristic column algorithm for solving it. The results show that the model and algorithm can solve large-scale crew scheduling problems in a short time, and at the same time improve the robustness of computer crew scheduling. Geng Yao established a multi-objective optimization model based on minimizing crew scheduling costs, shortest crew duty time, and maximum crew utilization, and solved it using a genetic algorithm, obtaining relatively satisfactory simulation results. However, with the changes in the regulations on crew duty time, flight time, and rest time in the Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules (CCAR-121-R7) promulgated by the Ministry of Transport on March 15, 2021, the traditional model for the construction of targets such as crew duty time, flight time, and scheduling costs can no longer meet the needs of enterprises. Therefore, it is necessary to build a model that complies with the latest regulations to meet the requirements of crew scheduling. In recent years, many scholars have added new constraints to the traditional crew scheduling model. For example, Quesnel et al. introduced the working language constraint in the crew rotation problem into the crew pairing problem. The study showed that with a slight increase in computational complexity, the scheduling rate that violates the language constraint can be reduced by 90%. Quesnel et al. also added a special constraint aimed at evenly distributing working time to all employees. [6]Santosa et al. proposed the Simple Iterative Alternative Method (SIMA) method to solve the airline crew scheduling problem in order to minimize the number of assigned crew members to cover all scheduled flights. From the experiments, the SIMA method produced better results in terms of roster quality and computation time. [7] Ma Hong and other scholars from Zhejiang University built a crew scheduling model based on the consistency of personnel work shifts and the consistency of personnel overnight cities, and used a heuristic algorithm to solve it. Through verification and calculation of actual data, the results significantly improved the consistency of the crew scheduling plan. [8] . However, no scholar has taken the operating cost, the optimal number of aircraft usage and the characteristics of the route as the optimization objectives at the same time, and the latest flight time, duty time, rest time and other conditions of the flight crew of large aircraft as constraints to schedule the flight crew. The present invention combines the latest management regulations on the duty time, flight time and rest time of the crew in the Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules (CCAR-121-R7) promulgated by the Ministry of Transport on March 15, 2021, and takes the operating cost, the number of aircraft usage and the characteristics of the route as the optimization objectives to construct a large aircraft flight crew scheduling model. Summary of the invention

[0003] In view of this, the present invention aims to provide a flight crew scheduling method for large aircraft public air transportation.

[0004] In order to optimize the flight crew scheduling, the flight crew scheduling model constructed by the present invention is divided into two parts, the first part is the aircraft allocation model, and the second part is the flight crew scheduling model.

[0005] Part I: Aircraft Allocation Model

[0006] The problem of aircraft task allocation is mainly solved by first setting constraints and then minimizing the number. Therefore, the mathematical model of this problem can be decomposed into two parts: one is to generate a flight task group that meets all regulatory constraints, and the other is to minimize the number of aircraft in this task group. The specific mathematical model is as follows:

[0007] X m =[a i a j a k a l ]i,j,k,l∈NN={1,2,……,N n}

[0008]

[0009]

[0010]

[0011]

[0012]

[0013]

[0014] Where: X m A task group for any flight;

[0015] N n Expressed as the total number of all known flight missions;

[0016] a i 、a j 、a k 、a l Represented as the original flight mission;

[0017] a tai It is represented as the landing time of the ith flight;

[0018] a taj It is represented as the landing time of the jth flight;

[0019] a tak It is represented as the landing time of the kth flight;

[0020] a tdj It is represented as the departure time of the jth flight;

[0021] a tdk It is represented as the departure time of the kth flight;

[0022] a tdl It is represented as the departure time of the lth flight;

[0023] a di Represents the landing airport code of the i-th flight;

[0024] a dj Represented as the landing airport code of the jth flight;

[0025] a dk Represented as the landing airport code of the kth flight;

[0026] a dl It is the landing airport code of the lth flight;

[0027] a ai Represents the departure airport code of the i-th flight;

[0028] a ajIt is represented as the departure airport code of the jth flight;

[0029] a ak Represented as the departure airport code of the kth flight;

[0030] a al It is the departure airport code of the lth flight;

[0031] Where tsxy represents the stopover limit time, x represents the stopover airport, where x=1 represents Beijing, Shanghai, Guangzhou, Shenzhen, Chengdu, Tianjin, Hangzhou, Chongqing, Kunming, Xi'an and other major international airports, and x=2 represents other airports. This part of the settings is based on the minimum stopover time regulations in the latest documents of the Civil Aviation Administration of China. If the minimum stopover time in the Civil Aviation Administration of China documents changes, this part of the settings can also be changed synchronously; y represents the number of seats.

[0032]

[0033] min∑C m X m

[0034]

[0035] C i ∈{0,1}

[0036] Where: m represents the flight task group; n represents the flight set that performs the task, C m The coefficient takes 1 or 0. When the flight task group does not meet the constraints, it takes 0; when the flight task meets the constraints, it takes 1; a mn It represents the constraint condition, that is, it takes 1 when a flight plan group corresponds to a flight that performs the task, and takes 0 otherwise;

[0037]

[0038] The above mathematical model is used to calculate the minimum number of flights and flight arrangements required each day. The flight mission allocation can be completed by arranging these output flight missions into the daily missions of the aircraft.

[0039] Part II: Flight Crew Scheduling Model

[0040] The main idea of ​​the flight crew scheduling model comes from the set partitioning problem [1]Based on the model of this problem, a mathematical model of various limiting factors for pilots that meets the requirements of CCAR-121-R7 regulations is made. Therefore, the mathematical model is divided into three parts to complete: the first is to obtain a flight task group that meets the requirements of the regulations on duty and flight time restrictions, the second is to minimize this task group to obtain all tasks, and the third is to obtain a task group arrangement across days based on all the tasks obtained, while meeting the rest time requirements of the regulations and the weekly and monthly duty and flight time restrictions. The specific mathematical model is as follows:

[0041] A={1,2,……,A a}

[0042] Y b =[r i r j r k r l ]i,j,k,l∈A

[0043] st zqb =t al -(t di -2.5)

[0044] st fxb =t al -t di

[0045]

[0046]

[0047]

[0048] A a It is expressed as the total number of all missions including make-up flight missions;

[0049] st zqb It is represented as the duty limit time of the bth task;

[0050] st fxb It is represented as the flight limit time of the bth task;

[0051] t al It represents the estimated arrival time of the lth flight for this mission;

[0052] t di represents the estimated departure time of the i-th flight for this mission;

[0053] Where, t fxwIndicates the flight restriction time. w represents the reporting time, where w=1 means the duty start time is 0:00-4:59, w=2 means the duty start time is 5:00-11:59, and w=3 means the duty start time is 12:00-23:59.

[0054]

[0055] Where, t zqef Indicates the duty limit time. e represents the duty start time. Where e=1 means the duty start time is 0:00-4:59, e=2 means the duty start time is 5:00-11:59, and e=3 means the duty start time is 12:00-23:59. f represents the number of flight segments

[0056]

[0057] Where, t rgh It represents the duty time limit of the expanded crew to meet the on-board rest level. g represents the number of pilots in the crew and h represents the on-board rest level.

[0058]

[0059] Y b Represents the flight task group that meets the constraint requirements;

[0060] r i 、r j 、r k 、r l It indicates that the flight mission of aircraft adjustment has been increased.

[0061] min∑C b Y b

[0062]

[0063] C b ∈{0,1}

[0064] Where: a represents the flight plan serial number; b represents the flight task group number; C b The coefficient takes 1 or 0. When the flight mission does not meet the constraint conditions, it takes 0; when the flight mission meets the constraint conditions, it takes 1; b ab represents the constraint condition, that is, when a flight plan corresponds to a flight task, it takes 1, and the rest takes 0; B represents the set of flight task group numbers, and the task table with the minimum number of crews required for the total number of days is obtained.

[0065]

[0066] y e=[Y1 Y2 Y3 … Y i ]

[0067] st∑ty ei ≤t fl

[0068] Y a1 =Y d2

[0069] Y a2 =Y d3

[0070] Y a3 =Y d4

[0071] …

[0072] Y ae-1 =Y de

[0073] y e Represented as any i-day flight task group

[0074] Y1, Y2...Y i Any item on the i-th day of the task group that meets the requirements for a total of i days;

[0075] ty ej It is represented as the duty time of the jth task;

[0076] Where, t fl Indicates the duty limit time. According to the regulations, the number of days is seven days. fl =60;

[0077] Y a1 It represents the landing airport at the end of the first mission;

[0078] Y a2 It indicates the landing airport at the end of the second mission;

[0079] Y a3 It indicates the landing airport at the end of the 3rd mission;

[0080] Y ae-1 It is represented as the landing airport at the end of the e-1th mission;

[0081] Y d2 It represents the departure airport at the start of the second mission;

[0082] Y d3 It represents the departure airport at the beginning of the third mission;

[0083] Y d4 It indicates the departure airport at the beginning of the 4th mission;

[0084] Y de It represents the departure airport at the beginning of the e-th mission;

[0085] min∑C e y e

[0086]

[0087] C e ∈{0,1}

[0088] Where: a represents the sequence number of the flight plan; e represents the number of the task group on day i; C e The coefficient takes 1 or 0. When the flight mission on day i does not meet the constraint conditions, it takes 0; when the flight mission on day i meets the constraint conditions, it takes 1; b ae It means that when the flight plan sequence number a corresponds to the number e of the task group on the i-th day, it takes 1, and otherwise takes 0; E represents the set of numbers of the task group on the i-th day;

[0089]

[0090] Compared with the prior art, the flight crew scheduling method for large aircraft public air transport described in the present invention has the following advantages:

[0091] The model constructed by the present invention is divided into two parts, an aircraft allocation model and a flight crew scheduling model. The aircraft allocation model can improve the utilization rate of aircraft by setting long-distance route priorities, that is, the same aircraft can be allocated more flight missions than the actual ones and achieve a longer total flight time, and the length of the total flight time will also reduce the cost index CI, thereby reducing the total cost of the airline. In the flight crew scheduling model, by setting bases with a larger proportion of pilots as priorities, the complexity of changing aircraft models due to insufficient number of pilots can be effectively reduced, and the problem of excessive cost of dispatching pilots due to insufficient number of pilots can be reduced. DETAILED DESCRIPTION

[0092] It should be noted that, in the absence of conflict, the embodiments of the present invention and the features in the embodiments may be combined with each other.

[0093] In the description of the present invention, it should be noted that, unless otherwise clearly specified and limited, the terms "installed", "connected", and "connected" should be understood in a broad sense, for example, it can be a fixed connection, a detachable connection, or an integral connection; it can be a mechanical connection or an electrical connection; it can be a direct connection, or it can be indirectly connected through an intermediate medium, or it can be the internal communication of two components. For ordinary technicians in this field, the specific meanings of the above terms in the present invention can be understood by specific circumstances.

[0094] 1. Data Collection

[0095] According to the constructed flight crew scheduling model, the flight schedule is at least a flight schedule of at least seven consecutive calendar days for a certain aircraft type or multiple aircraft types of the airline. According to Articles 121.487 and 121.493 of the "Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules", the flight duty period limit of seven consecutive calendar days for pilots is clearly given. In order to meet the regulatory requirements and verify the rationality of the airline crew scheduling model, at least seven consecutive calendar days of crew scheduling plans are required, and the flight tasks of general airlines are cycled on a weekly basis. For the above reasons and drawing on the practical experience of airline crew scheduling, a 7-day crew scheduling is used here to verify the reliability of the model. According to Article 121.485, different onboard rest levels have different duty limit times, and the onboard rest facilities of a certain aircraft type or multiple aircraft types must also be given. The minimum stopover time of a certain type or multiple types of aircraft can be obtained by consulting the "Civil Aviation Normality Statistics Method" to obtain the minimum stopover time of a certain type or multiple types of aircraft at each airport, so as to carry out single-day or multi-day flight connection work.

[0096] Through the collation of the above required information, and the reference to relevant laws and regulations and the actual data of a company, the following contents were obtained:

[0097] 1. The flight schedule for the 2019 winter and spring season provided by a company’s operations control center;

[0098] 2. Through personal experience with the B787-9 aircraft, I learned from inquiries that the rest area on this type of aircraft has an independent, lie-flat rest area;

[0099] 3. Minimum stopover time for all aircraft types at major airports.

[0100] 4. Relevant restrictions on pilots’ flight duty period, flight time and rest time in the “Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules”.

[0101] Table 1: A company's B787-9 flight schedule for the winter and spring of 2019

[0102]

[0103]

[0104] Note: In the table, D represents domestic flights and I represents international flights. The times are Beijing time. 907 / 908 (L) is the PEK-MAD-GRU route; 879 / 880 (L) is the PEK-YUL-HAV route.

[0105] Table 2 Minimum stop time

[0106]

[0107] 2. Data collation

[0108] According to the data obtained, the flight schedule is modified into a schedule for independent flight missions for each day. After the production is completed, it is compiled with a serial number, and the take-off and landing airports are changed to digital codes for subsequent program reading. The company's B787-9 aircraft has a total of 293 seats. The minimum stopover time at Beijing, Hongqiao, Pudong, Guangzhou, Shenzhen, Tianjin, Hangzhou, Chongqing, Kunming, Xi'an airports and international flights is determined to be 75 minutes, and 65 minutes at other airports. According to Chapter P of Part 121, regarding the crew's flight duty period, flight restriction time and rest time, the flight and duty restriction time tables for this data are produced, as shown in Tables 3 and 4:

[0109] Table 3 Pilot flight limit time

[0110]

[0111] Table 4 Pilot duty limit time

[0112]

[0113] The "Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules" clearly stipulates that the flight duty period for pilots for seven consecutive calendar days is 60 hours. Since the data involves many long-distance routes, according to the relevant provisions of the regulations, if there is a time difference of 6 or more hours between the time zone where the flight duty period ends and the time zone where the crew members' base is located, when the crew members return to the base, the certificate holder must arrange a rest period of at least 48 consecutive hours for them. Specific embodiment:

[0115] The tool chosen for model verification is MATLAB. The entire mathematical model and specific data are combined through MATLAB programming to obtain the final result. The program is divided into three files: aircraft flight mission schedule (flight.m), pilot mission group (crew.m) and pilot crew scheduling (fxcw_result.m).

[0116] First, for the flight task plan, after importing the original data and stopover time, the task group of the flight plan is generated. The constraints that need to be met are: first, the landing airport of the current flight and the departure airport of the next flight on a single day must be the same; second, the minimum stopover time limit of the landing airport of the current flight must be met. Through these two constraints, the single-day aircraft flight plan task group that meets the above requirements is obtained.

[0117] Through the integer programming function intlinprog function learned in the course, the objective function calculated is to minimize the number of aircraft required per day, that is, to select the minimum number of task groups. The program is as follows:

[0118]

[0119]

[0120] f is the coefficient matrix of the objective function; m is the number of aircraft flight planning task groups for the day; A and b are the coefficient matrices in the inequality constraint; Aeq and beq are the coefficient matrices in the equality constraint. The sum function is used in the equality constraint to make the flight numbers appear only once in all task groups, so as to achieve the minimum number of aircraft in the program; lb and ub are the upper and lower limits of the independent variables, and the value here is m1, which is the value with the longest number of task groups for the day.

[0121] Since there are tasks that require aircraft transfer between days, it is impossible to directly use the program to perform programming calculations, so manual sorting tasks are performed. Manual sorting is arranged according to the constraints of the same take-off and landing airports and the minimum stopover time limit. In principle, the aircraft task allocation work is carried out for 7 consecutive days in the order of "first arrive, first leave". The aircraft transfer flight tasks that appear are written in the flight data, a new complete version of the flight schedule is generated, and the results are displayed as the aircraft task allocation results for seven consecutive days with the flight serial number.

[0122] Next is the pilot task group. After importing the new version of the complete flight schedule and the single-day pilot limit time after data processing, the first step is to calculate the flight time according to the take-off and landing time of each flight segment. The second step is to calculate the duty start time. The company manual stipulates that the general flight duty period is 2-2.5 hours in advance from the initial flight task to the end of the last flight. The block time is the actual duty period. Therefore, the duty start time of this model is 2.5 hours in advance according to the take-off time of the first flight segment. Since the B787-9 aircraft has a large number of international routes, there are many routes where the single mission time exceeds the upper limit of the two-person crew's duty time. Therefore, the duty time is greater than 9 hours. The separation is used as an expansion crew, and the remaining tasks continue to be handled by the double crew. The order of continuous tasks obtained by aircraft task allocation generates a task matrix that requires at least 2, 3, and 4 flight crew members. Since the task of the expansion crew is a single task, it cannot be split any further. Only the non-expanded crew needs to be split. After splitting the tasks of the non-expanded crew, all possible task situations are obtained, and then it is judged whether the flight duty period of this task meets the requirements. After eliminating the unsatisfied tasks, we can get a task group that meets the daily time limit specified in the regulations. The number of times each flight task appears in these task groups is 1, which is a complete pilot task group. Finally, the integer programming function intlinprog function is used for calculation, and the program is as follows:

[0123]

[0124] The main body of the program is similar to the aircraft task allocation, and the objective function is the minimum value of the flight task to be executed for all flights. Through this calculation, the number of pilot task groups and the specific flight number of the task group for seven consecutive days of this data are obtained, and the day information, take-off and landing time of the flight segment, and duty phase time are written into the matrix to complete the task.

[0125] The last program is an assignment of the task group to each complete crew, which is similar to the pilot model, so only the pilot model needs to be verified. First, after entering the results obtained from the previous program and the new flight schedule, the legal rest time limit of 6 hours of time difference is implemented, and the tasks with a time difference of more than 6 hours are marked. The task group results obtained by the previous program are split and spliced. Since the upper limit of the pilot's duty for 7 consecutive calendar days is 60 hours, that is, the maximum work in 7 consecutive calendar days will not exceed 3 days, so in addition to the single-day task, a multi-day splicing task is required. This splicing task can be divided into 2-day and 3-day splicing. After the program completes the two-day splicing, it is found that for this flight schedule, there is no task group that has work for two consecutive days. Therefore, when performing three-day splicing, there is no need to splice one by one. Only the tasks that meet the interval of more than one day need to be spliced, which reduces a lot of workload. The above-mentioned completed task groups are solved by integer programming. Using the intlinprog function to calculate, the objective function is the minimum value of the flight task that all flights must be executed. Using 0-1 variables, if the task is selected, it is 1, and if it is not selected, it is 0. The constraint condition is that the number of occurrences of all flight numbers on that day is 1, and the program calculation is performed. The main body of the program is the same as the task group method above, and the final number of required pilot crews and the specific arrangement of the tasks can be obtained.

[0126] According to the results of the MATLAB program calculation: First, regarding the aircraft allocation problem, the minimum number of aircraft required on the first and sixth days is 12, and 13 on the remaining days. Based on the 7-day scheduling results, at least 13 B787-9 aircraft are needed to complete all flight missions, of which 5 aircraft transfer flights are required. Secondly, the number of pilot task groups is 152, which means that the order of aircraft task allocation needs to be decomposed into 152 task groups that meet the pilot's single-day requirements. Finally, the number of flight crews is 122, which represents the number of flight crews that meet the 7-day continuous duty time limit and rest time requirements in this given flight plan.

[0127] Table 5 Pilot task groups that meet the aircraft task allocation sequence

[0128]

[0129]

[0130]

[0131] Table 6 Flight crew scheduling results

[0132]

[0133] Note: The tasks obtained in this table are flight tasks that meet the pilot flight time limit, flight duty period and rest time requirements on days 1-7. Each number represents a task for a crew. The final total demand for pilots can be calculated based on the number of people required for each task.

[0134] According to the company's existing data, the company has a total of 15 B787-9 aircraft, of which 2 cannot perform missions due to engine replacement. It is completely feasible to use 13 aircraft to perform missions in actual operation. The 7-day schedule is the result of manual scheduling, taking into account the time priority order and the minimum stopover time limit. According to the actual flight schedule of the company's B787-9 aircraft from December 2 to December 8 of a certain year on the website, the statistical comparison with this data is shown in Table 7:

[0135] Table 7 Comparison of flight plan data before and after application of the model

[0136]

[0137] By comparing the programming data with the company's actual flight operation data for this week, it can be seen that this mathematical model has performed more flight missions in total, which has greatly improved the utilization rate of aircraft and fully utilized the time of each aircraft. Cost index CI = (costs related to flight hours ÷ flight hours) / fuel price. By analyzing the formula, when the costs related to flight hours and fuel prices do not change, the greater the number of flight hours, the smaller the CI cost index, and the smaller the reduction in fuel consumption. It can be seen that the airline aircraft task allocation cost index produced by this mathematical model is smaller, which saves the company's operating costs. Moreover, among the effective flight plan tasks executed by flights, long-distance international flight tasks account for a high proportion. For the operation of the company's large wide-body aircraft, the scheduling plan of this model is more advantageous.

[0138] There are 152 flight mission groups and 122 pilot crews. The long-distance international route company requires two captains and two co-pilots. The Beijing-Madrid-Sao Paulo route and the Beijing-Montreal-Havana route must have a backup group at the intermediate stopover airports of Madrid and Montreal. Flight missions that only perform Beijing-Madrid and Beijing-Montreal flights do not require a backup group. According to the data, there are 72 long-distance missions that require a four-crew configuration. In order to meet the company's special requirements, a total of 196 pilots with captain qualifications and 196 co-pilots holding flight licenses for this type of aircraft are required. The number of pilots given by the company and the comparison with the required number are shown in Table 8:

[0139] Table 8 Comparison of pilot data

[0140]

[0141] In order to perform all the tasks in the flight schedule, 196 pilots with B787-9 captain qualification and 196 pilots with co-pilot qualification are needed. The company has a B787-9 flight squadron at Beijing and Shenzhen terminals. From the data, it is impossible to use all the personnel of these two flight squadrons. By analyzing the flight plan, it can be seen that there is a Beijing-Shanghai route, which can use a small amount of resources from the Shanghai flight squadron. At the same time, Tianjin and Beijing are close, and a small number of pilots can also be borrowed from the Tianjin flight squadron to complete the task. In actual operation, according to the description of the company's staff, the pilots of the B787-9 flight squadron are often insufficient, and pilots will be borrowed from other flight squadrons, or directly replaced with A330-300 aircraft to perform some long-distance missions with 4 crew members. According to the known data, this model can effectively reduce the complex problem of changing models due to insufficient number of pilots, and at the same time reduce the problem of excessive cost of dispatching pilots due to insufficient number of pilots. Make full use of the pilots in the headquarters and Shenzhen B787-9 flight squadron to reduce dependence on other aircraft models and flight squadrons at other bases.

[0142] To sum up, the flight crew scheduling model constructed by the present invention takes the latest management regulations on crew duty time, flight time, and rest time in the Large Aircraft Public Air Transport Carrier Operation Qualification Certification Rules (CCAR-121-R7) promulgated by the Ministry of Transport on March 15, 2021 as constraints. The constructed model complies with the management regulations of airlines on flight crew duty time, flight time and rest time after the implementation of the new regulations.

[0143] The flight crew scheduling model constructed by the present invention is divided into two parts: an aircraft allocation model and a flight crew allocation model. By setting the aircraft allocation priority through the aircraft allocation model, the utilization rate of aircraft can be effectively improved. For example, by giving priority to long routes (international routes), the same number of aircraft can achieve more flight missions and a longer total flight time than actually allocated. The increase in total flight time also reduces the airline cost index CI, reducing the total cost of the airline.

[0144] The flight crew scheduling model constructed by the present invention can take the airline base and other bases with a large number of flight crews as the priority for flight crew allocation, thereby effectively reducing the complex problem of changing aircraft models due to insufficient number of pilots, and at the same time reducing the problem of excessively high cost of dispatching pilots due to insufficient number of pilots.

[0145] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.

Claims

1. A flight crew scheduling method for large aircraft public air transport, characterized by: include Taking operating costs, aircraft usage and route characteristics as optimization targets, and crew duty time, flight time and rest time as constraints, a large aircraft flight crew scheduling model is constructed; The large aircraft flight crew scheduling model includes an aircraft allocation model and a flight crew scheduling model; The aircraft allocation model consists of two parts: one is to generate a flight task group that meets all regulatory constraints, and the other is to minimize the number of aircraft in this task group. The specific mathematical model is as follows X m =[ai aj ak al]i,j,k,l∈NN={1,2,……,N n } Where: X m A task group for any flight; N n Expressed as the total number of all known flight missions; a i 、a j 、a k、 a l Represented as the original flight mission; a tai It is represented as the landing time of the ith flight; a taj It is represented as the landing time of the jth flight; a tak It is represented as the landing time of the kth flight; a tdj It is represented as the departure time of the jth flight; a tdk It is represented as the departure time of the kth flight; a tdl It is represented as the departure time of the lth flight; a di Represents the landing airport code of the i-th flight; a dj Represented as the landing airport code of the jth flight; a dk Represented as the landing airport code of the kth flight; a dl It is the landing airport code of the lth flight; a ai Represents the departure airport code of the i-th flight; a aj It is represented as the departure airport code of the jth flight; a ak Represents the departure airport code of the kth flight; a al It is the departure airport code of the lth flight; Where t sxy It represents the stopover limit time, x represents the stopover airport, where x=1 represents Beijing, Shanghai, Guangzhou, Shenzhen, Chengdu, Tianjin, Hangzhou, Chongqing, Kunming, Xi'an and other major international airports, and x=2 represents other airports; y represents the number of seats; min∑C m X m Where: m represents the flight task group; n represents the flight set that performs the task, C m The coefficient takes 1 or 0. When the flight task group does not meet the constraints, it takes 0; when the flight task meets the constraints, it takes 1; a mn represents the constraint condition, that is, when a flight plan group corresponds to a flight that performs a task, it takes 1, and otherwise takes 0; M represents the set of flight task group numbers, and N represents the flight sequence number; The above mathematical model is used to calculate the minimum number of flights and flight arrangements required each day. The flight mission allocation can be completed by arranging these output flight missions into the daily missions of the aircraft. The flight crew scheduling model includes three parts: first, obtaining a flight task group that meets the legal duty and flight time limit requirements; second, minimizing this task group to obtain all tasks; third, based on all the obtained tasks, under the condition of meeting the legal rest time requirements and weekly and monthly duty and flight time limits, obtaining a task group arrangement across days. The specific mathematical model is as follows: A={1,2,……,A a } Y b =[r i r j r k r l ]i,j,k,l∈A st zqb =t al -(t di -2.5) st fxb =t al -t di A a It is expressed as the total number of all missions including make-up flight missions; st zqb It is represented as the duty limit time of the bth task; st fxb It is represented as the flight limit time of the bth task; t al It represents the estimated arrival time of the lth flight for this mission; t di represents the estimated departure time of the i-th flight for this mission; Where, t fxw It is expressed as the flight restriction time, w represents the reporting time, where w=1 means the duty start time is 0:00-4:59, w=2 means the duty start time is 5:00-11:59, and w=3 means the duty start time is 12:00-23:59; Where, t zqef Indicates the duty limit time, e represents the duty start time, where e=1 means the duty start time is 0:00-4:59, e=2 means the duty start time is 5:00-11:59, e=3 means the duty start time is 12:00-23:59, and f represents the number of flight segments Where, t rgh It represents the duty time limit of the expanded crew to meet the on-board rest level, g represents the number of pilots in the crew, and h represents the on-board rest level; Y b Represents the flight task group that meets the constraint requirements; r i 、r j 、r k 、r l It indicates that the flight mission of adding aircraft adjustment is increased; min∑C b AND b C b ∈{0,1} Where: a represents the serial number of the flight plan; b represents the number of the flight task group; C b The coefficient takes 1 or 0. When the flight mission does not meet the constraint conditions, it takes 0; when the flight mission meets the constraint conditions, it takes 1; b ab represents the constraint condition, that is, when a flight plan corresponds to a flight task, it takes 1, and the rest takes 0; A represents the set of flight plans for the i-th day, and B represents the set of flight task group numbers, and the task table with the minimum number of crews required for the total number of days is obtained; <h2 style=";text-align:left;direction:ltr">y<h2 style=";text-align:left;direction:ltr"> e <h2 style=";text-align:left;direction:ltr"> (Y1 Y2 Y3……Y<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> ] s.t.∑ty ei ≤t fl AND a1 =And d2 AND a2 =And d3 AND a3 =And d4 …… AND ae-1 =And de y e Represented as any i-day flight task group Y1, Y2...Y i Any item on the i-th day of the task group that meets the requirements for a total of i days; ty ej It is represented as the duty time of the jth task; Where, t fl Indicates the duty limit time. According to the regulations, the number of days is seven days. fl =60; Y a1 It represents the landing airport at the end of the first mission; Y a2 It indicates the landing airport at the end of the second mission; Y a3 It indicates the landing airport at the end of the 3rd mission; Y ae-1 It is represented as the landing airport at the end of the e-1th mission; Y d2 It represents the departure airport at the start of the second mission; Y d3 It represents the departure airport at the beginning of the third mission; Y d4 It indicates the departure airport at the beginning of the 4th mission; Y de It represents the departure airport at the beginning of the e-th mission; min∑C e and e C e ∈{0,1} Where: a represents the sequence number of the flight plan; e represents the number of the task group on day i; C e The coefficient takes 1 or 0. When the flight mission on day i does not meet the constraint conditions, it takes 0; when the flight mission on day i meets the constraint conditions, it takes 1; b ae It means that when the flight plan sequence number a corresponds to the number e of the task group on the i-th day, it takes 1, and otherwise takes 0; A represents the set of flight plans for the i-th day, and E represents the set of numbers of the task group on the i-th day; The flight schedule is derived through the above model; By solving the model, the flight crew scheduling method is obtained.

2. A flight crew scheduling device for large aircraft public air transport, characterized in that: include The model building device constructs a large aircraft flight crew scheduling model by taking operating costs, aircraft usage and route characteristics as optimization targets and crew duty time, flight time and rest time as constraints; The large aircraft flight crew scheduling model includes an aircraft allocation model and a flight crew scheduling model; The aircraft allocation model consists of two parts: one is to generate a flight task group that meets all regulatory constraints, and the other is to minimize the number of aircraft in this task group. The specific mathematical model is as follows X m =[a i a j a k a l ]i,j,k,l∈N N={1,2,……,N n } Where: X m A task group for any flight; N n Expressed as the total number of all known flight missions; a i 、a j 、a k、 a l Represented as the original flight mission; a tai It is represented as the landing time of the ith flight; a taj It is represented as the landing time of the jth flight; a tak It is represented as the landing time of the kth flight; a tdj It is represented as the departure time of the jth flight; a tdk It is represented as the departure time of the kth flight; a tdl It is represented as the departure time of the lth flight; a di Represents the landing airport code of the i-th flight; a dj Represented as the landing airport code of the jth flight; a dk Represented as the landing airport code of the kth flight; a dl It is the landing airport code of the lth flight; a ai Represents the departure airport code of the i-th flight; a aj It is represented as the departure airport code of the jth flight; a ak Represents the departure airport code of the kth flight; a al It is the departure airport code of the lth flight; Where t sxy It represents the stopover limit time, x represents the stopover airport, where x=1 represents Beijing, Shanghai, Guangzhou, Shenzhen, Chengdu, Tianjin, Hangzhou, Chongqing, Kunming, Xi'an and other major international airports, and x=2 represents other airports; y represents the number of seats; min∑C m X m Where: m represents the flight task group; n represents the flight set that performs the task, C m The coefficient takes 1 or 0. When the flight task group does not meet the constraint conditions, it takes 0; when the flight task meets the constraint conditions, it takes 1; a mn represents the constraint condition, that is, when a flight plan group corresponds to a flight that performs a task, it takes 1, and otherwise takes 0; M represents the set of flight task group numbers, and N represents the flight sequence number; The above mathematical model is used to calculate the minimum number of flights and flight arrangements required each day. The flight mission allocation can be completed by arranging these output flight missions into the daily missions of the aircraft. The flight crew scheduling model includes three parts: first, obtaining a flight task group that meets the legal duty and flight time limit requirements; second, minimizing this task group to obtain all tasks; third, based on all the obtained tasks, under the condition of meeting the legal rest time requirements and weekly and monthly duty and flight time limits, obtaining a task group arrangement across days. The specific mathematical model is as follows: A={1,2,……,A a } Y b =[r i r j r k r l ]i,j,k,l∈A st zqb =t al -(t di -2.5) st fxb =t al -t di A a It is expressed as the total number of all missions including make-up flight missions; st zqb It is represented as the duty limit time of the bth task; st fxb It is represented as the flight limit time of the bth task; t al It represents the estimated arrival time of the lth flight for this mission; t di represents the estimated departure time of the i-th flight for this mission; Where, t fxw It is expressed as the flight restriction time, w represents the reporting time, where w=1 means the duty start time is 0:00-4:59, w=2 means the duty start time is 5:00-11:59, and w=3 means the duty start time is 12:00-23:59; Where, t zqef Indicates the duty limit time, e represents the duty start time, where e=1 means the duty start time is 0:00-4:59, e=2 means the duty start time is 5:00-11:59, e=3 means the duty start time is 12:00-23:59, and f represents the number of flight segments Where, t rgh It represents the duty time limit of the expanded crew to meet the on-board rest level, g represents the number of pilots in the crew, and h represents the on-board rest level; Y b Represents the flight task group that meets the constraint requirements; r i 、r j 、r k 、r l It indicates that the flight mission of adding aircraft adjustment is increased; min∑C b AND b C b ∈{0,1} Where: a represents the serial number of the flight plan; b represents the number of the flight task group; C b The coefficient takes 1 or 0. When the flight mission does not meet the constraint conditions, it takes 0; when the flight mission meets the constraint conditions, it takes 1; b ab represents the constraint condition, that is, when a flight plan corresponds to a flight task, it takes 1, and the rest takes 0; A represents the set of flight plans for the i-th day, B represents the set of flight task group numbers, Obtain a task list with the minimum number of crews required for the total number of days; <h2 style=";text-align:left;direction:ltr">y<h2 style=";text-align:left;direction:ltr"> e <h2 style=";text-align:left;direction:ltr"> (Y1 Y2 Y3……Y<h2 style=";text-align:left;direction:ltr"> i <h2 style=";text-align:left;direction:ltr"> ] s.t.∑ty ei ≤t fl AND a1 =And d2 AND a2 =And d3 AND a3 =And d4 …… AND ae-1 =And de y e Represented as any i-day flight task group Y1, Y2...Y i Any item on the i-th day of the task group that meets the requirements for a total of i days; ty ej It is represented as the duty time of the jth task; Where, t fl Indicates the duty limit time. According to the regulations, the number of days is seven days. fl =60; Y a1 It represents the landing airport at the end of the first mission; Y a2 It indicates the landing airport at the end of the second mission; Y a3 It indicates the landing airport at the end of the 3rd mission; Y ae-1 It is represented as the landing airport at the end of the e-1th mission; Y d2 It represents the departure airport at the start of the second mission; Y d3 It represents the departure airport at the beginning of the third mission; Y d4 It indicates the departure airport at the beginning of the 4th mission; Y de It represents the departure airport at the beginning of the e-th mission; min∑C e and e C e ∈{0,1} Where: a represents the sequence number of the flight plan; e represents the number of the task group on day i; C e The coefficient takes 1 or 0. When the flight mission on day i does not meet the constraint conditions, it takes 0; when the flight mission on day i meets the constraint conditions, it takes 1; b ae It means that when the flight plan sequence number a corresponds to the number e of the task group on the i-th day, it takes 1, and otherwise takes 0; A represents the set of flight plans for the i-th day, and E represents the set of numbers of the task group on the i-th day; The flight schedule is derived through the above model; A solving device obtains a flight crew scheduling method by solving the model.

3. An electronic device, characterized in that: include at least one processor, and at least one memory in communication with the processor, wherein: The memory stores program instructions that can be executed by the processor, and the processor calls the program instructions to execute the method according to claim 1. 4 . A non-volatile computer-readable storage medium, when the computer-executable instructions are executed by one or more processors, causes the processors to perform the method of claim 1 .

Images