Method, system and device for selecting optimal flight market demand value prediction model, storage medium and program product

Abstract

The invention belongs to the technical field of air transportation, and provides a method, system and device for selecting an optimal flight market demand value prediction model, a storage medium and a program product, and the method comprises the steps: obtaining all flight information of a specified airline, and sorting and storing the flight information; acquiring a historical flight market demand value based on the flight information; establishing a historical flight data pool based on the historical flight market demand value; the method comprises the following steps: dividing a historical flight data pool into a Base Data Pool and an Iterative Data Pool; and predicting historical flight samples in the historical flight data pool by using the market demand value prediction model, calculating and comparing accumulative errors according to a prediction result, and determining an optimal market demand value prediction model. According to the invention, the efficiency of airline income management work is improved, the function of the income management system is optimized, and the overall performance of the income management system is enhanced.

Description

Technical Field

[0001] The present invention belongs to the field of aviation transportation technology, and in particular relates to a method, system, device, storage medium and program product for selecting an optimal flight market demand value prediction model. Background Art

[0002] Revenue management is the management of prices and seats by airlines using scientific means such as forecasting and optimization, so that every seat on every segment of every flight can be sold to different types of passengers at different prices in a timely manner, thereby obtaining maximum revenue.

[0003] The Revenue Management System (Basic Edition) is a system that automatically manages the inventory of flights that have not yet departed based on forecasting and optimization models, using flight schedules, inventory, departures and freight rate data. The forecasting module of the revenue management system uses a variety of forecasting models to forecast key values ​​such as market demand values ​​to select the optimal forecasting model to meet the output uniqueness requirements. Among them, determining that a certain undeparted forecast target flight is the market demand value of the future flight departure DCP or each DCP is an important output of the forecasting subsystem of the revenue management system. Although the revenue management system uses a variety of forecasting models for forecasting, how to select the optimal forecasting model has become a problem and difficulty that needs to be solved. Summary of the invention

[0004] To solve the above problems, the present invention provides a method and system for selecting an optimal market demand value prediction model.

[0005] A method for selecting an optimal flight market demand value prediction model of the present invention comprises:

[0006] Obtain all flight information of a designated airline and organize the flight information into a database; based on the flight information, obtain historical flight market demand values;

[0007] Based on the historical flight market demand values, a historical flight data pool is established;

[0008] Dividing the historical flight data pool into a base data pool and an iterative data pool;

[0009] The market demand value prediction model is used to predict the historical flight samples in the historical flight data pool, and the cumulative errors are calculated and compared according to the prediction results to determine the optimal market demand value prediction model.

[0010] Furthermore, the flight information includes full flight information and flight incremental information;

[0011] The full flight information includes flight schedule data SCH, flight booking data INV, and flight operation data FLT in the ASTD aviation standard data of the lCS flight control system: airline, flight number, origin, destination, flight departure date and time, flight arrival date and time, electronic ticket identification, cabin structure table, flight physical layout number and the maximum number of seats available for sale on the flight.

[0012] Furthermore, the flight information is the flight information of a designated airline and a designated flight segment, with the current date or the system date of the revenue management system as the basis for the next year and the past three years.

[0013] Further, the market demand value prediction model is used to predict the historical flight samples in the historical flight data pool, and the cumulative errors are calculated and compared according to the prediction results to determine the optimal market demand value prediction model, which specifically includes:

[0014] S4-1. Based on the historical flight samples in the Base Data Pool, the market demand value prediction model is used to predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline. 23 , denoted as ModelOutput(i);

[0015] S4-2, based on the historical flight samples in the iterative update data pool, using the market demand value prediction model, predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline 23 , denoted as DemandDcp 23 (i);

[0016] The single error SimpleError and the cumulative error AccumulatedError of the historical flight samples in each market demand value prediction model are calculated, and the iterative process is:

[0017] SimpleError(i)=ModelOutput(i)-DemandDcp 23 (i);

[0018] AccumulatedError(i)=AccumulatedError(i)+SimpleError(i) 2 ;

[0019] S4-3, deleting the historical flight samples in step S4-2 from the iterative data pool IterativeData Pool and moving them to the base data pool Base Data Pool;

[0020] S4-4, determining an updating method of the historical flight data pool according to the number of historical flight samples in the historical flight data pool;

[0021] S4-5, determine whether the iterative data pool is empty; if it is not empty, repeat steps S4-1 to S4-4; if it is empty, execute step S4-6;

[0022] S4-6. By comparing the accumulated error value AccumulatedError of each market demand value prediction model, determine the market demand value prediction model with the smallest accumulated error value as the optimal market demand value prediction model.

[0023] Furthermore, the method for updating the historical flight data pool includes:

[0024] Fixed Base Data Pool technology and methods and Increasing Base Data Pool technology and methods.

[0025] Furthermore, the minimum historical flight sample threshold of the base data pool Base Data Pool must be greater than or equal to 5, and the maximum number of cycles must not be less than 15.

[0026] The present invention also provides a system for selecting an optimal flight market demand value prediction model, the system comprising:

[0027] An acquisition module is used to acquire all flight information of a designated airline and organize the flight information into a database; based on the flight information, acquire historical market demand values;

[0028] An establishment module is used to establish a historical flight data pool based on the historical flight market demand value;

[0029] A division module, used for dividing the historical flight data pool into a base data pool and an iterative data pool;

[0030] The module for determining the optimal market demand value prediction model is used to use the market demand value prediction model to predict the historical flight samples in the historical flight data pool, calculate and compare the cumulative errors based on the prediction results, and determine the optimal market demand value prediction model.

[0031] The present invention also provides a device, including a processor, which is coupled to a memory; the processor is used to read and execute the computer program stored in the memory to implement the aforementioned method of selecting an optimal market demand value prediction model.

[0032] The present invention also provides a computer-readable storage medium storing a program or instruction. When the program or instruction is executed on a computer, the computer executes the aforementioned method of selecting an optimal market demand value prediction model.

[0033] The present invention also provides a computer program product, including a computer program or instructions, characterized in that when the computer program or instructions are executed by a processor, the aforementioned method of selecting an optimal market demand value prediction model is implemented.

[0034] Compared with the prior art, the present invention has the following advantages:

[0035] The present invention significantly improves the work efficiency of airlines by optimizing the prediction model selection, while enhancing the functionality and performance of the revenue management system. By accurately predicting the market demand from flight segments to cabin levels, the prediction model selection problem is solved, filling the gap in the systematic operation in this field in China, and bringing more accurate market positioning and higher revenue potential to airlines. BRIEF DESCRIPTION OF THE DRAWINGS

[0036] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0037] Figure 1 A schematic flow chart of a method for selecting an optimal market demand value prediction model according to the present invention;

[0038] Figure 2 A schematic diagram of the system structure of the present invention for selecting an optimal flight market demand value prediction model;

[0039] Figure 3 It is a structural schematic diagram of the electronic device of the present invention;

[0040] Figure 4 Schematic diagram of the iterative process of an embodiment of the present invention. DETAILED DESCRIPTION

[0041] In order to make the purpose, technical solution and advantages of the embodiments of the present invention clearer, the technical solution in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0042] The following are some professional terms for explaining the present invention:

[0043] Revenue Management System (Basic Edition): A system that automatically manages the inventory of flights that have not departed based on forecasting and optimization models using flight schedules, inventory, departure and freight rate data.

[0044] Inventory: refers to the information that affects flight seat sales decisions, including seat availability (Availability), seat reservation value (SeatSold), number of available seats (Seat Open), and various control parameters (Inventory Parameter).

[0045] ICS (Inventory Control System): The airline reservation system mentioned in this article is also called inventory control system.

[0046] Data Collection Points (Dcp): set by airlines based on their own route quality, flight attributes and other considerations. The data collection code is determined by the number of days before departure and has a one-to-one correspondence with the number of days before departure. Usually, airlines set Dcp to 24, and the closer to the flight departure date, the larger the Dcp value. For example: the data collection point Dcp is called departure Dcp and the corresponding number of days before departure is 0 days.

[0047] Fixed-DCP: A data collection point set by an airline, which is determined by the number of days until departure. An airline can set up two sets of data collection points in the flight control system (or revenue management system), one for domestic flight routes and one for international flight routes.

[0048] Floating-DCP: A data collection point that is not a fixed data collection point. For example, a flight is scheduled to leave the port in 20 days, but the DCP 11 The corresponding distance to the port is 24 days, DCP 12 The corresponding distance to the port is 18 days, so this data collection point is a floating data collection point and can be recorded as DCP 11.5 .

[0049] Final-DCP: The data collection point on the day of flight departure is also called departure DCP;

[0050] Latest data collection point Last actual-DCP: the data collection point for the most recent data acquisition of flights that have not departed; the revenue management forecasting and optimization module uses the data of the flight segment-level inventory acquired based on the Last actual-DCP as the independent variable.

[0051] Booking Value (Bkg): The space or weight of a seat, sleeping space, baggage or cargo reserved by a passenger and pre-allocated to the passenger.

[0052] Market demand value (Demand): Within a booking period and under certain price and service prerequisites, the number of flight products that passengers are willing and able to purchase at such a price or service.

[0053] Base Data Pool: A flight group consisting of historical (sample) flights with departure dates farther away than the future flights, usually historical (sample) flights with departure dates more than one year before the future flights. For example, if the departure date of a future flight is July 8, 2024, the historical (sample) flights in this data pool are historical (sample) flights with departure dates earlier than July 7, 2023. The (sample) data of historical (sample) flights include but are not limited to market demand value Demand and booking value Bkg.

[0054] Iterative Data Pool: Historical (sample) flights that do not fall into the Base Data Pool can be understood as remaining recent samples and fall into the iterative data pool Base Data Pool.

[0055] In one embodiment of the present invention, a method for selecting an optimal market demand value prediction model is provided, such as Figure 1 As shown, the method includes:

[0056] S1. According to the two-letter code of the designated airline, all flight information of the designated airline is obtained, and the flight information is sorted and stored in the database;

[0057] The flight information includes full flight information and flight incremental information;

[0058] The complete flight information includes flight schedule data SCH, flight reservation data INV, and flight operation data FLT in the ASTD aviation standard data of the ICS flight control system: airline, flight number, origin, destination, flight departure date and time, flight arrival date and time, electronic ticket identification, cabin structure table, flight physical layout number and the maximum number of seats available for sale on the flight.

[0059] The flight information is the flight information for the specified flight segments of a specified airline, based on the current date or the system date of the revenue management system for the next year and the past three years.

[0060] S2. Establishing a historical flight data pool based on the historical flight market demand value;

[0061] S3, dividing the historical flight data pool into a base data pool and an iterative data pool;

[0062] S4. Predicting the historical flight samples in the historical flight data pool using the market demand value prediction model, calculating and comparing the cumulative errors based on the prediction results, and determining the optimal market demand value prediction model;

[0063] S4-1. Based on the historical flight samples in the Base Data Pool, the market demand value prediction model is used to predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline. 23 (or the predetermined value BkgDcp 23 ), denoted as ModelOutput(i);

[0064] S4-2, based on the historical flight samples in the iterative update data pool, using the market demand value prediction model, predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline 23 (or the predetermined value BkgDcp 23 ), denoted as DemandDcp 23 (i);

[0065] Calculate the single error SimpleError and the cumulative error AccumulatedError of the historical flight sample in each market demand value prediction model;

[0066] Among them, Figure 4 The schematic diagram of the iterative process is shown below. The iterative process is:

[0067] SimpleError(i)=ModelOutput(i)-DemandDcp 23 (i);

[0068] AccumulatedError(i)=AccumulatedError(i)+SimpleError(i) 2 .

[0069] S4-3, deleting the historical flight samples in step S4-2 from the iterative data pool IterativeData Pool and moving them to the base data pool Base Data Pool;

[0070] S4-4. Determine a method for updating the historical flight data pool according to the number of historical flight samples in the historical flight data pool, specifically including:

[0071] The updating method of the historical flight data pool includes:

[0072] Fixed Base Data Pool technology and method means that each time a historical flight sample in the Iterative Data Pool is added to the Base Data Pool, the earliest historical flight sample in the Base Data Pool will be deleted, so that the total number of samples in the Base Data Pool remains unchanged;

[0073] Increasing Base Data Pool technology and methods means that each time a historical flight sample in the Iterative Data Pool is added to the Base Data Pool, the original historical flight sample in the Base Data Pool is not deleted; that is, the number of historical flight samples in the Base Data Pool in each cycle is always one more than the previous cycle;

[0074] Specifically, when the number of historical flight samples is greater than or equal to 50, the Fixed Base Data Pool technology and method is used; when the number of historical flight samples is less than 50, the Increasing Base Data Pool technology and method is used;

[0075] For example: if the number of historical flight samples is greater than 60, the Fixed BaseData Pool technology and method is used;

[0076] If the number of historical flight samples is greater than 40 and less than 60, a fixed base data pool BaseData Pool technology and method is used, and 10 samples less than the number of historical flight samples are selected to form a base data pool BaseData Pool;

[0077] If the number of historical flight samples is less than 40, the increasing base data pool BaseData Pool technology and method is adopted.

[0078] S4-5, determine whether the iterative data pool is empty; if it is not empty, repeat steps S4-1 to S4-4; if it is empty, execute step S4-6;

[0079] Among them, the minimum historical flight sample threshold of the Base Data Pool must be greater than or equal to 5; the maximum number of cycles of this method is not less than 15;

[0080] If the number of historical flight samples in the historical flight data pool is greater than the maximum number of cycles plus the minimum number of historical flight samples in the base data pool, then the maximum number of cycles of this method will be the maximum number of cycles, and the number of historical flight samples in the base data pool will be equal to the number of historical flight samples minus the maximum number of cycles;

[0081] If the number of historical flight samples in the historical flight data pool is less than the maximum number of cycles plus the minimum number of historical flight samples in the base data pool BaseData Pool, then the minimum number of historical flight samples in the base data pool Base Data Pool of this method is equal to its threshold, and the maximum number of cycles is the number of historical flight samples minus the number of historical flight samples in the base data pool Base DataPool.

[0082] S4-6. By comparing the accumulated error value AccumulatedError of each market demand value prediction model, determine the market demand value prediction model with the smallest accumulated error value as the optimal market demand value prediction model.

[0083] The embodiment of the present invention also provides a system for selecting an optimal market demand value prediction model, such as Figure 2 As shown, the system comprises:

[0084] The acquisition module 201 is used to acquire all flight information of a designated airline and organize the flight information into a database; based on the flight information, acquire the historical market demand value;

[0085] Establishing module 202, used for establishing a historical flight data pool based on the historical flight market demand value;

[0086] A division module 203, used for dividing the historical flight data pool into a base data pool and an iterative data pool;

[0087] The module 204 for determining the optimal market demand value prediction model is used to predict the historical flight samples in the historical flight data pool using the market demand value prediction model, calculate and compare the cumulative errors according to the prediction results, and determine the optimal market demand value prediction model.

[0088] like Figure 3 As shown, an embodiment of the present invention also provides a device, including: a processor 301, the processor 301 is coupled to a memory 302, and the processor 301 is used to read and execute a computer program stored in the memory 302 to implement a method for selecting an optimal market demand value prediction model as described in the above method embodiment.

[0089] An embodiment of the present invention also provides a computer-readable storage medium, which stores a program or instruction. When the above program or instruction runs on a computer, the computer executes a method for selecting an optimal market demand value prediction model as described in the above method embodiment.

[0090] The present invention also provides a computer program product, including a computer program or instructions, which, when executed by a processor, implements the aforementioned method for selecting an optimal market demand value prediction model.

[0091] Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent substitutions for some of the technical features therein; and these modifications or substitutions do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for selecting an optimal market demand value prediction model, characterized in that: The method comprises: Obtain all flight information of a designated airline and organize the flight information into a database; based on the flight information, obtain historical flight market demand values; Based on the historical flight market demand values, a historical flight data pool is established; Dividing the historical flight data pool into a base data pool and an iterative update data pool; The market demand value prediction model is used to predict the historical flight samples in the historical flight data pool, and the cumulative errors are calculated and compared according to the prediction results to determine the optimal market demand value prediction model.

2. The method according to claim 1, characterized in that: The flight information includes full flight information and flight incremental information; The complete flight information includes flight schedule data SCH, flight reservation data INV, and flight operation data FLT in the ASTD aviation standard data of the ICS flight control system: airline, flight number, origin, destination, flight departure date and time, flight arrival date and time, electronic ticket identification, cabin structure table, flight physical layout number and the maximum number of seats available for sale on the flight.

3. The method according to claim 1, characterized in that The flight information is the flight information for the specified flight segments of a specified airline, based on the current date or the system date of the revenue management system for the next year and the past three years.

4. The method according to claim 1, characterized in that The method of using the market demand value prediction model to predict the historical flight samples in the historical flight data pool, calculating and comparing the cumulative errors according to the prediction results, and determining the optimal market demand value prediction model specifically includes: S4-1. Based on the historical flight samples in the Base Data Pool, the market demand value prediction model is used to predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline. 23 , denoted as ModelOutput(i); S4-2, based on the historical flight samples in the iteratively updated data pool, using the market demand value prediction model, predict and output the market demand value DemandDcp of the departure Dcp of the specified flight segment of the specified airline 23 , denoted as DemandDcp 23 (i); The single error SimpleError and the cumulative error AccumulatedError of the historical flight samples in each market demand value prediction model are calculated, and the iterative process is: SimpleError(i)=ModelOutput(i)-DemandDcp 23 (i); AccumulatedError(i)=AccumulatedError(i)+SimpleError(i) 2 ; S4-3, deleting the historical flight samples in step S4-2 from the iterative data pool Iterative DataPool and moving them to the base data pool Base Data Pool; S4-4, determining an updating method of the historical flight data pool according to the number of historical flight samples in the historical flight data pool; S4-5, determine whether the iterative data pool is empty; if it is not empty, repeat steps S4-1 to S4-4; if it is empty, execute step S4-6; S4-6. By comparing the accumulated error value AccumulatedError of each market demand value prediction model, determine the market demand value prediction model with the smallest accumulated error value as the optimal market demand value prediction model.

5. The method according to claim 1, characterized in that: The updating method of the historical flight data pool includes: Fixed Base Data Pool technology and method and Increasing Base Data Pool technology and method.

6. The method according to claim 1, characterized in that The minimum historical flight sample threshold of the Base Data Pool must be greater than or equal to 5, and the maximum number of cycles must not be less than 15.

7. A system for selecting an optimal market demand value prediction model, characterized in that: The system comprises: An acquisition module is used to acquire all flight information of a designated airline and organize the flight information into a database; based on the flight information, acquire historical market demand values; An establishment module is used to establish a historical flight data pool based on the historical flight market demand value; A division module, used for dividing the historical flight data pool into a base data pool and an iterative data pool; The module for determining the optimal market demand value prediction model is used to use the market demand value prediction model to predict the historical flight samples in the historical flight data pool, calculate and compare the cumulative errors based on the prediction results, and determine the optimal market demand value prediction model.

8. A device, characterized in that comprising a processor coupled to a memory; The processor is used to read and execute the computer program stored in the memory to implement a method for selecting an optimal market demand value prediction model as described in any one of claims 1-6.

9. A computer-readable storage medium, characterized in that: A program or instruction is stored, and when the program or instruction is executed on a computer, the computer is caused to execute a method for selecting an optimal market demand value prediction model as described in any one of claims 1-6.

10. A computer program product comprising a computer program or instructions, characterized in that When the computer program or instruction is executed by a processor, a method for selecting an optimal market demand value prediction model as described in any one of claims 1-6 is implemented.

Images