A method and device for generating an intelligent flight crew ring

Abstract

The purpose of this application is to provide a method and device for generating an intelligent flight crew ring. This application searches for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew to obtain a duty day set; constructs a flight ring network diagram according to the duty day set; constructs a flight ring model based on the set of flights to be scheduled and the initial task ring set; uses the branch pricing method to solve the main problem of the flight ring model to obtain a feasible task ring; uses a multi-label shortest path algorithm to find the optimal path for solving the main problem of the flight ring model to obtain an optimized task ring, solves the flight ring model according to the feasible task rings of each flight and the optimized task ring, and determines the ring scheme of the intelligent flight crew. Therefore, the quality and robustness of the scheme can be guaranteed within a limited solution time, and it has high research significance and application value.

Description

Technical Field

[0001] The present application relates to the field of computers, and in particular to a method and device for generating an intelligent flight crew ring. Background Art

[0002] Since the development of the aviation industry, the resource allocation of Chinese airlines is still in the form of manual allocation, and the quality of the scheduling plan also mainly depends on the rich experience of the person in charge. With the improvement of people's living standards, the demand for air travel has increased rapidly, the fleet size has expanded rapidly, and the route network has become increasingly complex. The Statistical Bulletin on the Development of the Civil Aviation Industry in 2020 released by the Civil Aviation Administration of China in 2021 shows that in 2020, the country's passenger airlines carried out a total of 3.5206 million flights, including 3.1164 million normal flights, with an average flight regularity rate of 88.52%. The development of routes and the expansion of fleet size are still continuing. As of the end of 2020, my country has a total of 64 transport airlines, 3,903 transport aircraft, and 5,581 scheduled flight routes.

[0003] The flight crew usually refers to all pilots involved in flying a flight, with a minimum configuration of two people, namely the captain and the co-pilot. When the flight time or duty time is long, the flight crew will be expanded, that is, the number of flight crew members exceeds the minimum number required by the aircraft model to operate the aircraft, so that other qualified flight crew members can replace a flight crew member, and the replaced flight crew member can rest during the flight.

[0004] According to the regulations of the Civil Aviation Administration, any crew member must arrange a continuous 48-hour rest period within 144 hours before performing a flight mission or primary backup. In order to reduce the difficulty of problem solving and to ensure that the crew members have a 48-hour rest at their base, the crew scheduling problem is usually divided into two sub-problems: ring formation and assignment. The ring formation stage will complete the series connection of all flights, backups and other tasks to be scheduled during the planning period, and obtain multiple task strings that start from the base and return to the base, with a duration of no more than 96 hours, that is, generally no more than 5 calendar days. The flight hour fee of the crew, the crew overnight allowance, the cost of rides or flights, etc. mainly depend on the construction process of the task ring. In addition, the flight time margin, flight duty time margin, rest time margin, etc. related to the robustness of the plan also need to be adjusted in the ring formation stage.

[0005] For the crew ring formation problem, in addition to the increase in business volume, the continuous updating and refinement of the work rules of the Civil Aviation Administration's 121 regulations and the personalized business needs supplemented by various airlines have also led to a higher complexity in problem solving. At present, the main implementation solution is manual ring formation, which is mainly based on the flight plan and route structure, with the primary premise of meeting the compliance of the mission ring, adding necessary ride or flight positions, and connecting all flights to be scheduled within the planning period into a mission ring set. In such a large and complex aviation network, facing the complicated scheduling rules, the scheduling and allocation of resource dispatchers are gradually unable to cope with it, and the importance of informatization and automation is becoming increasingly prominent. On the one hand, the quality of the ring formation solution reflects the overall utilization efficiency of crew resources. On the other hand, its robustness also reflects to a certain extent the possibility of flight chain delays caused by crew reasons during actual operation. Summary of the invention

[0006] One purpose of the present application is to provide a method and device for generating an intelligent flight crew ring assembly, so as to solve the problem that the ring assembly business in the prior art is completely carried out manually, which is time-consuming and labor-intensive, and is prone to errors, and it is impossible to optimize the crew ring assembly globally, which has obvious limitations.

[0007] According to one aspect of the present application, a method for generating an intelligent flight crew ring is provided, the method comprising:

[0008] Searching for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew to obtain a duty day set, wherein the rule conditions include the Civil Aviation Administration's regulatory restrictions and the rules for group ring scene configuration;

[0009] Constructing a flight group ring network diagram according to the duty day set;

[0010] Constructing a flight group ring model based on the set of flights to be scheduled and the set of initial task rings, wherein the flight group ring model includes a model objective function and constraint conditions;

[0011] Using the branch-and-price method to solve the main problem of the flight group loop model, and obtain the feasible task loop of each flight;

[0012] A multi-label shortest path algorithm is used to find the optimal path for solving the main problem of the flight group ring model, and an optimized task ring is obtained. The flight group ring model is solved according to the feasible task rings of each flight and the optimized task ring, and the group ring scheme of the intelligent flight crew is determined.

[0013] Optionally, all duty days that meet the rule conditions are searched out based on the basic data of the intelligent flight crew, including:

[0014] Obtain basic data related to intelligent flight crew scheduling;

[0015] The solution preference of the ring combination scheme is determined according to the basic data and the rules of the ring combination scenario are configured.

[0016] Optionally, constructing a flight group ring network diagram according to the duty day set includes the following steps:

[0017] Determine the planned start date and the planned end date of each personnel base and each personnel task according to the duty day set;

[0018] For any personnel base, according to the maximum number of calendar days of the mission loop set in the scenario configuration, for each day from the planned start date to the planned end date, a spatiotemporal network diagram with a duration of one to the maximum number of calendar days of the mission loop is constructed;

[0019] According to the screening rules, the target duty days that need to be added to the space-time network diagram are screened out from the duty day set, and connecting arcs are set between the duty days that meet the location connection, rest rules and scene configuration rules in the Civil Aviation Administration regulations to complete the construction of the flight group ring network diagram.

[0020] Optionally, the screening rule comprises the following steps:

[0021] If the date corresponding to the start time of the duty day is equal to the start date of the space-time network diagram, then continue to determine whether the starting airport of the duty day is the corresponding personnel base, and if so, add it to the space-time network diagram;

[0022] If the date of the duty day is equal to any middle date in the said space-time network diagram, it will be directly added;

[0023] If the date corresponding to the end time of the duty day is equal to the end date of the space-time network diagram, then continue to determine whether the arrival airport of the duty day is the corresponding personnel base, and if so, join it.

[0024] Optionally, constructing a flight group ring model based on the set of flights to be scheduled and the set of initial task rings includes:

[0025] Determine the set of flights to be scheduled and the set of initial mission rings, and set the thresholds of fixed integer variables;

[0026] Constructing a model objective function of the flight ring combination model according to the cost of each task ring in the initial task ring set, the decision variable selected by the ring combination scheme for the task ring, and the set of all task rings generated during the planning period;

[0027] The constraints of the flight group ring model are determined based on the constraints on set partitioning and the constraints on variable values.

[0028] Optionally, the model objective function represents minimizing the total cost of the ring combination solution, and the model objective function satisfies the following conditions:

[0029]

[0030] Among them, p represents the task ring, represents the cost of task ring p, is a decision variable of 0 or 1, indicating whether the task ring p is selected in the final ring combination scheme, and P represents the set of all generated task rings during the planning period.

[0031] Optionally, the constraint condition satisfies the following conditions:

[0032] ;

[0033] ;

[0034] Among them, L represents the set of flights to be scheduled, and l represents the planned flight segment.

[0035] Optionally, the flight group ring model is solved using a branching and pricing method, wherein the main problem solving includes the following steps:

[0036] Step 1, determine whether the current solution is an integer solution, if not, proceed to step 2, otherwise exit and obtain a new flight ring solution;

[0037] Step 2: Set up the collection , represents the final ring combination solution; when When Add to collection ,in, represents the threshold value of a fixed integer variable;

[0038] Step 3: For ,set up The Nether , indicating that the task ring p is selected. If there are other task rings The same flight as the mission ring p , set the variable value corresponding to the task ring Upper Bound , indicating the task cycle be eliminated;

[0039] Step 4: Obtain the linear relaxation solution of the main problem of the flight group ring model.

[0040] Optionally, a multi-label shortest path algorithm is used to find an optimal path for solving the main problem of the flight group ring model to obtain an optimized task ring, including:

[0041] Using a multi-label shortest path algorithm to find an optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network graph;

[0042] If a set of task rings that are beneficial to the optimization of the main problem solution is found, then the set of task rings is added to the task ring set and the process goes to step 1;

[0043] When a task ring that is conducive to optimizing the solution of the main problem cannot be found, jump to the following steps: construct a flight group ring model corresponding to the flight group ring network diagram.

[0044] Optionally, using a multi-label short-circuit algorithm to find an optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network graph includes the following steps:

[0045] Step S1, initializing the labels of all nodes in the flight group ring network, wherein the labels include the duty day set and the total inspection number included in the current partial path;

[0046] Step S2, traverse each node in turn according to the topological sorting, and obtain the set of its subsequent nodes. If the traversal is completed, execute step S5, otherwise execute step S3;

[0047] Step S3, for any label in the node and any node in the set of subsequent nodes, perform rule verification according to the rule conditions, if the verification shows that the rule conditions are met, jump to step S4, if the verification shows that the rule conditions are not met, re-execute step S3;

[0048] Step S4, generating a new label according to the labels that meet the rule conditions and the content of the target node in the set of subsequent nodes that have been successfully verified, and performing a dominance judgment between the new label and all existing labels in the target node. If the verification number of any existing label in the target node is smaller than the new label, the new label is not retained. Otherwise, the labels in the target node that are dominated by the new label are deleted, and the new label is added to the label set of the target node.

[0049] Step S5, traverse all labels on the task ring end node in the flight group ring network. If the check number recorded in the label is less than 0, the task ring corresponding to the path recorded by the label is used as the optimized path for solving the main problem of the flight group ring model.

[0050] Optionally, the method comprises:

[0051] Statistical evaluation indicators are performed based on the ring formation scheme, wherein the evaluation indicators include the average flight hours on duty day, the total number of positioning times, the total number of aircraft changes, the total number of stopovers by the flight captain, the total number of ring formations for consecutive morning shifts, the total number of ring formations for consecutive evening shifts, the crew cost, and the number of temporary overnight stops for the flight crew.

[0052] Optionally, the group ring scenario configuration includes the maximum number of continuous calendar days for the mission ring, the number of consecutive flight calendar days for the mission ring, the number of consecutive positioning times, the number of entry and exit times during the flight duty period, the morning shift definition settings, the evening shift definition settings, the flight connection type settings, the airport non-grouping restrictions and the base requirements.

[0053] According to another aspect of the present application, there is also provided a device for generating an intelligent flight crew ring, the device comprising:

[0054] one or more processors; and

[0055] A memory storing computer readable instructions, which, when executed, cause the processor to perform the operations of the method as described above.

[0056] According to another aspect of the present application, a computer-readable medium is provided, on which computer-readable instructions are stored. The computer-readable instructions can be executed by a processor to implement the method described above.

[0057] Compared with the prior art, the present application searches for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew to obtain a duty day set, wherein the rule conditions include the Civil Aviation Administration's regulatory restrictions and the rules for the ring grouping scenario configuration; constructs a flight ring grouping network diagram according to the duty day set; constructs a flight ring grouping model based on the set of flights to be scheduled and the initial task ring set, wherein the flight ring grouping model includes a model objective function and constraints; uses the branch pricing method to solve the main problem of the flight ring grouping model to obtain a feasible task ring for each flight; uses a multi-label shortest path algorithm to find the optimal path for solving the main problem of the flight ring grouping model to obtain an optimized task ring, solves the flight ring grouping model according to the feasible task rings of each flight and the optimized task ring, and determines the ring grouping scheme of the intelligent flight crew. . Thus, the optimal solution or a better solution to the crew ring grouping problem can be obtained in a short time, and the quality and robustness of the scheme can be guaranteed within a limited solution time, which has high research significance and application value. BRIEF DESCRIPTION OF THE DRAWINGS

[0058] Other features, objects and advantages of the present application will become more apparent by reading the detailed description of non-limiting embodiments made with reference to the following drawings:

[0059] Figure 1A schematic diagram showing a flow chart of a method for generating an intelligent flight crew ring according to one aspect of the present application;

[0060] Figure 2 A schematic diagram showing the format of a portion of a table recording flight information in an embodiment of the present application;

[0061] Figure 3 A schematic diagram of a flight group ring network diagram for a duty day in an embodiment of the present application is shown;

[0062] Figure 4 A schematic diagram of the flight group ring branch pricing process in an embodiment of the present application is shown;

[0063] Figure 5 A schematic diagram showing an aircraft unit and a flight path in an embodiment of the present application is shown.

[0064] The same or similar reference numerals in the drawings represent the same or similar components. DETAILED DESCRIPTION

[0065] The present application is described in further detail below in conjunction with the accompanying drawings.

[0066] In a typical configuration of the present application, the terminal, the device of the service network and the trusted party all include one or more processors (eg, a central processing unit (CPU)), an input / output interface, a network interface and a memory.

[0067] Memory may include non-permanent storage in a computer-readable medium, in the form of random access memory (RAM) and / or non-volatile memory, such as read-only memory (ROM) or flash RAM. Memory is an example of a computer-readable medium.

[0068] Computer readable media include permanent and non-permanent, removable and non-removable media that can be used to store information by any method or technology. Information can be computer readable instructions, data structures, program modules or other data. Examples of computer storage media include, but are not limited to, Phase-Change RAM (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of random access memory (RAM), Read-Only Memory (ROM), Electrically Erasable Programmable Read-Only Memory (EEPROM), Flash memory or other memory technology, Compact Disc Read-Only Memory (CD-ROM), Digital Versatile Disk (DVD) or other optical storage, cassettes, tape disk storage or other magnetic storage devices or any other non-transmission media that can be used to store information that can be accessed by a computing device. According to the definition in this article, computer-readable media does not include non-transitory media such as modulated data signals and carrier waves.

[0069] Figure 1 A schematic flow chart of a method for generating an intelligent flight crew ring according to one aspect of the present application is shown, the method comprising: steps S11 to S15, wherein:

[0070] Step S11, search for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew, and obtain a duty day set, wherein the rule conditions include the Civil Aviation Administration's regulatory restrictions and the rules of the ring scene configuration; here, when generating the ring of the intelligent flight crew, first generate the duty day, and use the depth-first search algorithm to search for all duty days that meet the Civil Aviation Administration's regulatory restrictions and the custom rules of the ring scene configuration according to the basic data such as flights, aircraft, aircraft models, and stopover time rules, so as to obtain a duty day set. Among them, the Civil Aviation Administration's regulatory restrictions include: flight time upper limit, duty day time upper limit, rest time lower limit, etc., and the rules of the ring scene configuration include: continuous positioning is not allowed, the number of entry and exit times shall not exceed 2 times, two flights are not allowed to be connected in series, and two airports are not allowed to appear in the same duty day, etc.; based on the characteristics of all flights and airports included in the duty day, such as: flight take-off and landing time, airport check-in time, etc., the characteristic data such as flight time, duty time, check-in time, check-out time, etc. of each duty day are calculated. If the scenario is configured with morning shift, evening shift, etc., the corresponding attributes of each duty day can be determined based on the definition.

[0071] Step S12, constructing a flight group ring network diagram according to the duty day set; here, based on the generated duty day set, the flight group ring network diagram is constructed according to rules such as rest requirements, so as to convert the generation of the group ring into a sub-problem of finding the shortest path in the flight group ring network diagram.

[0072] Step S13, constructing a flight ring combination model based on the set of flights to be scheduled and the initial task ring set, wherein the flight ring combination model includes a model objective function and constraints; here, when determining the ring combination scheme, a flight ring combination model is constructed according to the initial task ring set and the set of flights to be scheduled of the airline, and the model is converted into a model, and the model is solved to obtain a feasible ring combination and an optimized ring combination.

[0073] Step S14, using the branch pricing method to solve the main problem of the flight group ring model to obtain a feasible task ring; here, when using the branch pricing method to solve the model, the flight group ring model is first relaxed into a linear programming model, called the main problem, and then the main problem is solved to obtain the dual variable values ​​of each flight, thereby obtaining the feasible task ring of each flight, that is, first find some feasible task rings that can be judged to meet the rule conditions.

[0074] Step S15, using a multi-label shortest path algorithm to find the optimal path for solving the main problem of the flight group ring model in the flight group ring network diagram, obtain the optimized task ring, solve the flight group ring model according to the feasible task rings of each flight and the optimized task ring, and determine the group ring scheme of the intelligent flight crew. Here, the multi-label shortest path algorithm is used to find the optimal path for solving the main problem of the flight group ring model, obtain the optimized task ring, and repeat this step until the optimized path cannot be obtained; in addition to solving the main problem of the model, it is also necessary to use the multi-label shortest path algorithm based on the constructed flight group ring network diagram to find a set of task rings that are conducive to the optimization of solving the main problem, so as to add the variables corresponding to these optimized task rings to the main problem model, which is the above-mentioned linear programming model. When the optimized task ring cannot be found, all variables contained in the main problem are judged by rules, and the task rings that meet the rule conditions are added to the feasible task ring set, and the group ring scheme of the intelligent flight crew is determined according to the task ring set finally obtained. In this way, the optimal solution or a better solution to the crew grouping problem can be obtained in a short time. The obtained grouping solution is not only compliant and meets the regulations of the Civil Aviation Administration and the actual business requirements of the airline, but also has practical availability. It can control the fatigue level of the overall operating crew and the balance of various tasks, improve the scientific nature of crew resource scheduling, and lay a good foundation for the subsequent assignment of crew members.

[0075] In some embodiments of the present application, basic data related to intelligent flight crew scheduling is obtained; the solution preference of the ring combination scheme is determined based on the basic data and the rules of the ring combination scenario are configured. Here, the data of the basic elements related to crew scheduling are obtained and converted into csv table data for use by subsequent models and algorithms, wherein the basic data includes aircraft information (aircraft number, aircraft model, rest level, etc.), aircraft model information (aircraft model, sub-aircraft model, long and short aircraft model, etc.), airport information (airport four codes, airport three codes, city name, airport name, international and domestic attributes, etc.), flight information (planned starting and ending airports, planned starting and ending times, operating aircraft, flight attributes, etc.) and location information (flight information of other modes of transportation or other airlines), etc., wherein the format of some tables recording flight information is as follows Figure 2As shown. The relevant parameters in the algorithm model are configured to adjust the solution preference of the desired crew ring scheme, which can be divided into two categories: penalty classification and restriction classification. The penalty classification includes penalty points that will cause the airline's operating cost to increase in the ring scheme, such as: the number of overnight stays at foreign stations, the number of foreign crews, the number of trains, etc.; penalty points that cause the crew's fatigue level to increase, such as: the number of secondary entry and exit, the number of consecutive morning and evening shifts, the number of morning and evening jumps, etc.; penalty points that cause the scheme's robustness to decrease, such as: the number of aircraft changes, the number of insufficient rest time margins, the number of late landing times for the fourth consecutive flight day, etc. The restriction category includes the minimum airport transit time requirement, the flight time on duty day and the flight duty time limit, the rest period requirement, etc. And configure the rules of the ring scenario to convert the regulatory requirements and the personalized requirements in the actual crew ring business scenario into specific models and settings in the algorithm logic.

[0076] Continuing with the above embodiment, the group ring scenario configuration includes the maximum number of continuous calendar days for the mission ring, the number of continuous flight calendar days for the mission ring, the number of continuous positioning, the number of entry and exit times during the flight duty period, the morning shift definition setting, the evening shift definition setting, the flight connection type setting, the airport non-series grouping restriction, and the base requirements. Here, the maximum number of continuous calendar days for the mission ring is set to include the maximum number of continuous calendar days allowed for positioning, flights, backup and other tasks; the number of continuous flight calendar days for the mission ring is set to set the maximum number of continuous calendar days allowed for the existence of flight missions; the number of continuous positioning is set to set the number of times the crew members are allowed to take the positioning continuously during a duty period, and when continuous positioning is not allowed, it will be set to 1; the number of entry and exit times during the flight duty period is set to set whether the pilot is allowed to exit the flight duty period for rest, and the allowed number of entry and exit times can also be set separately according to the actual situation of the airport; the morning shift definition setting can be set by limiting the time period of the duty start time and the duty end time. The morning shift can be defined by the time period, etc., so as to meet a certain degree of regularity in the work and rest when forming the ring; the night shift definition setting can define the night shift by limiting the time period of the duty start time and the time period of the duty end time, so as to meet a certain degree of regularity in the work and rest when forming the ring; the flight fixed connection / prohibited connection / breakable connection setting can set the flight combination that must be connected in series or the flight combination that is prohibited from being connected in series according to the experience of business personnel or actual operation requirements, as well as the flight combination that can break the set scheduling rules; the airport non-connection restriction prohibits certain airports from being connected in series on the same duty day; the base requirement is to limit the number of mission rings departing from certain bases every day, etc.

[0077] In some embodiments of the present application, in step S12, the planned start date and the planned end date of each personnel base and each personnel task are determined according to the duty day set; for any personnel base, according to the maximum number of calendar days for the task loop set in the scenario configuration, for each day from the planned start date to the planned end date, a space-time network diagram with a duration of one to the maximum number of calendar days for the task loop is constructed; according to the screening rules, the target duty days that need to be added to the space-time network diagram in the duty day set are screened out, and connecting arcs are set between the duty days that meet the location connection, rest rules and scenario configuration rules in the Civil Aviation Administration's regulations to complete the construction of the flight group ring network diagram. Here, the steps for constructing the spatiotemporal network diagram of the duty day are as follows: Step 1, let B be the set of personnel bases, index b, StartDate be the planned start date, and EndDate be the planned end date; Step 2, for any personnel base, according to the maximum number of calendar days for the task loop set in the scenario configuration, for each day from the planned start date to the planned end date, construct a spatiotemporal network diagram with a duration of one to the maximum number of calendar days for the task loop; Step 3, determine whether the duty day d needs to be added to the network diagram according to the screening rules; Step 4, set connection arcs between the duty days that meet the location connection, rest rules and custom rules in the scenario configuration, where each duty day network contains three types of nodes: "task loop start", "task loop end" and "connection point", as well as two types of arcs: "duty day arc" and "connection arc", such as Figure 3 shown.

[0078] Continuing with the above embodiment, the screening rule includes the following steps: if the corresponding date of the start time of the duty day is equal to the start date of the space-time network diagram, then continue to judge whether the starting airport of the duty day is the corresponding personnel base, and if so, add it to the space-time network diagram; if the date of the duty day is equal to any date in the middle of the space-time network diagram, then directly add it; if the corresponding date of the end time of the duty day is equal to the end date of the space-time network diagram, then continue to judge whether the arrival airport of the duty day is the corresponding personnel base, and if so, add it. Here, when judging whether the duty day d needs to be added to the network diagram according to the screening rule, if the corresponding date of the start time of the duty day is equal to the start date of the network diagram, then it is also necessary to judge whether its starting airport is the corresponding personnel base, and if so, add it; if the date of the duty day is equal to a date in the middle of the network diagram, then directly add it; if the corresponding date of the end time of the duty day is equal to the end date of the network diagram, then it is also necessary to judge whether its arrival airport is the corresponding personnel base, and if so, add it.

[0079] In some embodiments of the present application, in step S13, the set of flights to be scheduled and the set of initial task rings are determined, and the threshold of fixed integer variables is set; the model objective function of the flight ring grouping model is constructed according to the cost of each task ring in the initial task ring set, the decision variable selected by the ring grouping scheme for the task ring, and the set of all task rings generated during the planning period; the constraint conditions of the flight ring grouping model are determined according to the constraints on the set partitioning and the constraints on the variable values. Here, the branch pricing method is used to determine the objective function and constraint conditions of the model. Specifically, the initial task ring set is obtained by using a heuristic algorithm, and the threshold of fixed integer variables is set at the same time. The objective function is determined according to the cost of each task ring, the set of all task rings generated during the planning period, and the decision variable (0-1 variable) of whether the task ring is selected in the final ring grouping scheme, and the set of flights that need to be scheduled is determined. The set of flights to be scheduled is a set of planned flight segments, so that the constraint conditions of the model are performed using the set of planned flight segments.

[0080] Continuing with the above embodiment, the model objective function represents minimizing the total cost of the ring assembly solution, and the model objective function satisfies the following conditions: , where p represents the task ring, represents the cost of task ring p, is a decision variable of 0 or 1, indicating whether the task ring p is selected in the final ring combination scheme, and P represents the set of all generated task rings during the planning period. The cost of mission loop p includes two parts. One part is related to the flight mission composition contained in the mission loop, representing the salary cost that the airline needs to pay for the pilots who fly the mission loop. The other part reflects the quality of the mission loop and is related to the penalty classification in the parameter configuration. When the mission loop has a feature corresponding to a penalty point, the corresponding penalty point needs to be added to its cost. It is a 0-1 variable, 0 variable means that the task ring p is not selected in the final ring combination plan, and 1 variable means that the task ring p is selected in the final ring combination plan. The objective function represents the minimization of the total cost of the ring combination plan, which includes the total cost of the pilot salary that the airline needs to pay to ensure the flight missions of all flights, as well as the additional costs such as overnight stays at foreign stations and placement.

[0081] Continuing with the above embodiment, the constraint condition satisfies the following conditions:

[0082] ;

[0083] ;

[0084] Where L represents the set of flights to be scheduled, and l represents the planned flight segment in the set of flights to be scheduled. Here, the first set partition constraint ensures that each planned flight segment is covered exactly once, and the second variable value constraint determines the value range of each decision variable.

[0085] In some embodiments of the present application, Figure 4 As shown, the branch pricing method is used to solve the main problem of the flight group ring model, wherein the main problem solution includes the following steps: Step 1, determine whether the current solution is an integer solution, if not, execute Step 2, otherwise exit and obtain a new flight group ring solution; Step 2, set the set , represents the final ring combination solution; when When Add to collection ,in, represents the threshold of a fixed integer variable; Step 3: For ,set up The Nether , indicating that the task ring p is selected. If there are other task rings The same flight as the mission ring p , set the variable value corresponding to the task ring Upper Bound , indicating the task cycle Eliminated; Step 4: Obtain the linear relaxation solution of the main problem of the flight group ring model.

[0086] Next, the multi-label shortest path algorithm is used to find the optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network graph; if a set of task rings that are conducive to the optimization of the main problem solution is found, then the set of task rings is added to the task ring set and jump to step 4; if no task ring that is conducive to the optimization of the main problem solution is found, then jump to step 1. Here, continue to refer to Figure 4 The column generation algorithm finds a set of task loops that are conducive to solving the optimization of the main problem based on the flight group loop network diagram of the duty day. , add it to the task ring collection, , go to step 4. Otherwise, when such a task ring cannot be found, jump to step 1 and iteratively process the solution of the flight group ring model, that is, use the multi-label shortest path algorithm to find a set of task rings that are conducive to the optimization of the main problem solution based on the constructed flight group ring network diagram, so as to add the variables corresponding to these optimized task rings to the main problem model. When the optimized task ring cannot be found, all variables contained in the main problem are judged by rules, and the task rings that meet the rule conditions are added to the feasible task ring set.

[0087] Specifically, a multi-label shortest path algorithm is used to find an optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network diagram, including the following steps: Step S1, initializing the labels of all nodes in the flight group ring network, wherein the labels include the duty day set and the total inspection number contained in the current partial path. In addition, depending on the scenario configuration, it is sometimes necessary to add corresponding labels. For example, when the mission ring is set to have a maximum number of consecutive calendar days allowed for a flight mission, an item will be added to the label to record the number of consecutive flight calendar days accumulated in the current partial path, which is used for rule judgment when connecting the paths in series; Step S2, traverse each node in turn according to the topological sorting, and obtain the set of its subsequent nodes. If the traversal is completed, execute step S5, otherwise execute step S3, and each node , the subsequent node set is S(n); step S3, for any label in the node and any node in the set of subsequent nodes, perform rule verification according to the rule conditions. If the verification is to meet the rule conditions, jump to step S4. If the verification is not to meet the rule conditions, re-execute step S3, that is, for any label p in point i and any node j in the subsequent set S(n), perform rule verification to determine whether the partial path recorded in label k continues to connect node j in series, which violates the regulations of the Civil Aviation Administration or the custom rules set in the scene configuration; step S4, generate a new label according to the content of the target node in the set of labels that meet the rule conditions and the subsequent nodes that have been successfully verified, and perform a dominance judgment on the new label and all the existing labels in the target node. If the verification number of any label in the target node is less than the new label, the new label will not be retained. Otherwise, the target node The labels dominated by the new label in the existing labels in the node are deleted, and the new label is added to the label set of the target node, that is, a new label q is generated according to the label p and the content of node j, and the label q is identified and occupied by all the existing labels in node j. If the check number of any label n in node j is less than label q, it is considered that label q is dominated and does not need to be retained. Otherwise, the labels dominated by label q in the existing labels of node j are deleted, and label q is added to the label set of node j; step S5, traverse all labels on the end node of the task ring in the flight group ring network, if the check number recorded in the label is less than 0, the task ring corresponding to the path recorded by the label is used as the optimized path for solving the main problem of the flight group ring model, that is, the task ring variable corresponding to the path recorded by the label is considered to be able to optimize the solution of the main problem of the above model, and is therefore adopted.

[0088] In some embodiments of the present application, the method includes: performing statistical evaluation indicators according to the ring group scheme, wherein the evaluation indicators include the average flight hours on duty day, the total number of positioning, the total number of aircraft changes, the total number of flight long stopovers, the total number of ring groups that continuously perform morning shifts, and the total number of ring groups that continuously perform evening shifts, crew costs, and the number of temporary overnight stations for flight crews. Here, the mission ring information is determined according to the ring group scheme, including information such as flights and positioning that constitute the mission ring, and evaluation indicators (KPI statistics) are performed on flight coverage, the number of overnight stays at foreign stations, the number of positioning, the number of long stopovers, the number of aircraft changes, the number of morning and evening shifts, etc.; wherein, the average flight hours on duty day: the duty day includes flight tasks, non-private crew additions, duty, backup, positioning, and task arrangements (simulator training, simulator flight, training, official business, other, regular training, examinations, and daily training). The value of the average flight hours on duty day can directly reflect the level of crew utilization. Total number of placements: Too many placements will not only directly affect the airline's revenue or increase costs, but also affect the subsequent crew's job satisfaction. Total number of aircraft changes: Too many aircraft changes during a duty day will not only affect the crew's experience, but also reduce the robustness of the overall crew scheduling plan to a certain extent. In actual operation, if there is a flight plan interruption, the different action paths of the crew and the aircraft often lead to chain delays. For example: Figure 5As shown, Aircraft 1: Flight F1->F2, Aircraft 2: Flight F3->F4, Crew 1: Flight F1->F4, Crew 2: Flight F3->F2. When Flight F1 is delayed, Flight F2 operated by Aircraft 1 may also be delayed due to insufficient stopover time. Crew 1 also has insufficient stopover time for changing aircraft, so Flight F4 operated by Crew 1 cannot take off as originally planned. Total number of long stopovers: Although the second entry and exit allows crew members to get better rest conditions between flight segments, it means that in the middle of the duty day, they need to complete a series of processes such as signing out, taking transportation between the airport and the rest area, and signing in. The actual experience of the crew is not good, especially when the remaining rest time in the rest area is not long, which is not worth the loss. Total number of consecutive morning shifts and consecutive night shifts: Human fatigue is closely related to biological rhythms and human work and rest habits. Continuous morning or night shifts will cause pilots to feel less fatigue, resulting in a decrease in pilot job satisfaction. Overnight days, overnight subsidy costs generated by scheduling, and hotel accommodation costs: The overnight costs in the crew cost are directly related to the number of days spent overnight at an outstation, and the number of overnight airports also needs to be controlled. On the premise of meeting the scheduling requirements, fewer overnight airports are easier to manage, and it is also easier to adjust the crew when flight delays occur. Number of temporary overnight stations for flight crews: If new overnight stations are allowed, it will effectively increase the possibility of duty day structure and group connection, which will help increase the average daily flight time of pilots. However, the addition of overnight points requires additional communication and agreement with hotels for airlines, which may also cause an increase in overnight costs. Therefore, while maximizing existing KPIs such as average daily flight time, the number of new temporary overnight points should also be limited.

[0089] Through the above-mentioned method for generating an intelligent flight crew ring in the present application, the quality and robustness of the solution can be guaranteed within a limited solution time, which has high research significance and application value.

[0090] In addition, an embodiment of the present application further provides a computer-readable medium having computer-readable instructions stored thereon, and the computer-readable instructions can be executed by a processor to implement the aforementioned method for generating an intelligent flight crew ring.

[0091] In one embodiment of the present application, a device for generating an intelligent flight crew ring is also provided, the device comprising:

[0092] one or more processors; and

[0093] A memory storing computer readable instructions, which, when executed, cause the processor to perform the operations of the method as described above.

[0094] For example, the computer readable instructions, when executed, cause the one or more processors to:

[0095] Searching for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew to obtain a duty day set, wherein the rule conditions include the Civil Aviation Administration's regulatory restrictions and the rules for group ring scene configuration;

[0096] Constructing a flight group ring network diagram according to the duty day set;

[0097] Constructing a flight group ring model based on the set of flights to be scheduled and the set of initial task rings, wherein the flight group ring model includes a model objective function and constraint conditions;

[0098] Using the branch-and-price method to solve the main problem of the flight group loop model, and obtain the feasible task loop of each flight;

[0099] A multi-label shortest path algorithm is used to find the optimal path for solving the main problem of the flight group ring model, and an optimized task ring is obtained. The flight group ring model is solved according to the feasible task rings of each flight and the optimized task ring, and the group ring scheme of the intelligent flight crew is determined.

[0100] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, if these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalents, the present application is also intended to include these modifications and variations.

[0101] It should be noted that the present application can be implemented in software and / or a combination of software and hardware, for example, it can be implemented using an application-specific integrated circuit (ASIC), a general-purpose computer or any other similar hardware device. In one embodiment, the software program of the present application can be executed by a processor to implement the steps or functions described above. Similarly, the software program of the present application (including related data structures) can be stored in a computer-readable recording medium, for example, a RAM memory, a magnetic or optical drive or a floppy disk and the like. In addition, some steps or functions of the present application can be implemented using hardware, for example, as a circuit that cooperates with a processor to perform various steps or functions.

[0102] In addition, a part of the present application may be applied as a computer program product, such as a computer program instruction, which, when executed by a computer, can call or provide the method and / or technical solution according to the present application through the operation of the computer. The program instruction for calling the method of the present application may be stored in a fixed or removable recording medium, and / or transmitted through a data stream in a broadcast or other signal-bearing medium, and / or stored in a working memory of a computer device that runs according to the program instruction. Here, according to an embodiment of the present application, a device is included, the device including a memory for storing computer program instructions and a processor for executing program instructions, wherein, when the computer program instruction is executed by the processor, the device is triggered to run the method and / or technical solution based on the aforementioned multiple embodiments according to the present application.

[0103] It is obvious to those skilled in the art that the present application is not limited to the details of the above exemplary embodiments, and that the present application can be implemented in other specific forms without departing from the spirit or basic features of the present application. Therefore, from any point of view, the embodiments should be regarded as exemplary and non-restrictive, and the scope of the present application is defined by the appended claims rather than the above description, and it is intended that all changes falling within the meaning and scope of the equivalent elements of the claims are included in the present application. Any figure mark in the claims should not be regarded as limiting the claims involved. In addition, it is obvious that the word "comprising" does not exclude other units or steps, and the singular does not exclude the plural. The words first, second, etc. are used to indicate names, and do not indicate any particular order.

Claims

1. A method for generating an intelligent flight crew ring, characterized in that: The method comprises: Obtain basic data of the intelligent flight crew and convert the basic data into csv table data, wherein the basic data includes aircraft information, aircraft model information, airport information, flight information and position information; Searching for all duty days that meet the rule conditions according to the basic data of the intelligent flight crew to obtain a duty day set, wherein the rule conditions include the Civil Aviation Administration's regulatory restrictions and the rules for group ring scene configuration; Constructing a flight group ring network diagram according to the duty day set; Constructing a flight group ring model based on the set of flights to be scheduled and the set of initial task rings, wherein the flight group ring model includes a model objective function and constraint conditions; Using the branch-and-price method to solve the main problem of the flight group loop model, and obtain the feasible task loop of each flight; Using a multi-label shortest path algorithm to find an optimal path for solving the main problem of the flight group ring model, obtaining an optimized task ring, solving the flight group ring model according to the feasible task rings of each flight and the optimized task ring, and determining a group ring scheme for the intelligent flight crew; Wherein, constructing a flight group ring network diagram according to the duty day set includes the following steps: Determine the planned start date and the planned end date of each personnel base and each personnel task according to the duty day set; For any personnel base, according to the maximum number of calendar days of the mission loop set in the scenario configuration, for each day from the planned start date to the planned end date, a spatiotemporal network diagram with a duration of one to the maximum number of calendar days of the mission loop is constructed; Filter out the target duty days that need to be added to the space-time network diagram from the duty day set according to the screening rules, set connection arcs between the duty days that meet the location connection, rest rules and scene configuration rules in the regulations of the Civil Aviation Administration of China, so as to complete the construction of the flight group ring network diagram, wherein the flight group ring network diagram includes three types of nodes: the start of the task ring, the end of the task ring, and the connection point, and two types of arcs: the duty day arc and the connection arc.

2. The method according to claim 1, characterized in that: According to the basic data of the intelligent flight crew, all duty days that meet the rule conditions are searched, including: Obtain basic data related to intelligent flight crew scheduling; The solution preference of the ring combination scheme is determined according to the basic data and the rules of the ring combination scenario are configured.

3. The method according to claim 1, characterized in that The screening rule comprises the following steps: If the date corresponding to the start time of the duty day is equal to the start date of the space-time network diagram, then continue to determine whether the starting airport of the duty day is the corresponding personnel base, and if so, add it to the space-time network diagram; If the date of the duty day is equal to any middle date in the said space-time network diagram, it will be directly added; If the date corresponding to the end time of the duty day is equal to the end date of the space-time network diagram, then continue to determine whether the arrival airport of the duty day is the corresponding personnel base, and if so, join it.

4. The method according to claim 1, characterized in that: Constructing a flight group ring model based on the set of flights to be scheduled and the set of initial task rings Constructing a flight group ring model includes: Determine the set of flights to be scheduled and the set of initial mission rings, and set the thresholds of fixed integer variables; Constructing a model objective function of a flight ring formation model according to the cost of each task ring in the initial task ring set, the decision variable selected by the ring formation scheme for the task ring, and all task ring sets generated during the planning period; The constraints of the flight group ring model are determined based on the constraints on set partitioning and the constraints on variable values.

5. The method according to claim 4, characterized in that The model objective function represents the total cost of minimizing the ring combination solution, and the model objective function satisfies the following conditions: Among them, p represents the task ring, c p represents the cost of task ring p, x p is a decision variable of 0 or 1, indicating whether the task ring p is selected in the final ring combination scheme, and P represents the set of all generated task rings during the planning period.

6. The method according to claim 5, characterized in that The constraints satisfy the following conditions: Among them, L represents the set of flights to be scheduled, and l represents the planned flight segment.

7. The method according to claim 4, characterized in that The main problem of the flight group ring model is solved by using the branch pricing method, wherein the main problem solving includes the following steps: Step 1, determine whether the current solution is an integer solution, if not, proceed to step 2, otherwise exit and obtain a new flight ring solution; Step 2: Set up the collection Indicates the final ring combination solution; when x is satisfied p ≥κ1, the corresponding x p Add to collection Among them, κ1 represents the threshold of the fixed integer variable; Step 3: For Setting x p Lower bound LB(x p )=1, indicating that the task ring p is selected. If there is another task ring p′ that contains the same flight l as the task ring p, set the variable value x corresponding to the task ring p′ , the upper bound UB(x p′ )=0, indicating that task ring p′ is eliminated; Step 4: Obtain the linear relaxation solution of the main problem of the flight group ring model.

8. The method according to claim 7, characterized in that The multi-label shortest path algorithm is used to find the optimal path for solving the main problem of the flight group ring model, and the optimized task ring is obtained, including: Using a multi-label shortest path algorithm to find an optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network graph; If a set of task rings that are beneficial to the optimization of the main problem solution is found, then the set of task rings is added to the task ring set, and the process goes to step 4; When no task loop that is conducive to optimizing the solution of the main problem can be found, jump to step 1.

9. The method according to claim 8, characterized in that Using a multi-label shortest path algorithm to find an optimal path for solving the main problem of the flight group ring model based on the constructed flight group ring network graph includes the following steps: Step S1, initializing the labels of all nodes in the flight group ring network, wherein the labels include the duty day set and the total inspection number included in the current partial path; Step S2, traverse each node in turn according to the topological sorting, and obtain the set of its subsequent nodes. If the traversal is completed, execute step S5, otherwise execute step S3; Step S3, for any label in the node and any node in the set of subsequent nodes, perform rule verification according to the rule conditions, if the verification shows that the rule conditions are met, jump to step S4, if the verification shows that the rule conditions are not met, re-execute step S3; Step S4, generating a new label according to the labels that meet the rule conditions and the content of the target node in the set of subsequent nodes that have been successfully verified, and performing a dominance judgment between the new label and all existing labels in the target node. If the verification number of any existing label in the target node is smaller than the new label, the new label is not retained. Otherwise, the labels in the target node that are dominated by the new label are deleted, and the new label is added to the label set of the target node. Step S5, traverse all labels on the task ring end node in the flight group ring network. If the check number recorded in the label is less than 0, the task ring corresponding to the path recorded by the label is used as the optimized path for solving the main problem of the flight group ring model.

10. The method according to claim 1, characterized in that The method comprises: Statistical evaluation indicators are performed based on the ring formation scheme, wherein the evaluation indicators include the average flight hours on duty day, the total number of positioning times, the total number of aircraft changes, the total number of stopovers by the flight captain, the total number of ring formations for consecutive morning shifts, the total number of ring formations for consecutive evening shifts, the crew cost, and the number of temporary overnight stops for the flight crew.

11. The method according to claim 1, characterized in that The ring group scenario configuration includes the maximum number of continuous calendar days for the mission ring, the number of continuous flight calendar days for the mission ring, the number of consecutive positioning times, the number of entry and exit times during the flight duty period, the morning shift definition settings, the evening shift definition settings, the flight connection type settings, the airport non-series grouping restrictions and the base requirements.

12. A device for generating an intelligent flight crew ring, characterized in that: The device comprises: one or more processors; and A memory storing computer readable instructions which, when executed, cause the processor to perform the operations of the method of any one of claims 1 to 11.

13. A computer-readable medium having computer-readable instructions stored thereon, wherein the computer-readable instructions can be executed by a processor to implement the method according to any one of claims 1 to 11.

Images