Method and system for origin-destination passenger demand forecast inference

• US 6,804,658 B2
• Filed: 12/06/2002
• Issued: 10/12/2004
• Est. Priority Date: 12/14/2001
• Status: Active Grant

• Current Assignee: Delta Air Lines Incorporated
• Original Assignee: Delta Air Lines Incorporated
• Inventors: Lim, Alvin, David, Andrew, Blankenbaker, John
• Primary Examiner(s): MEINECKE DIAZ, SUSANNA M
• Application Number:
  US10/313,556
• Publication Number:
  US 20030115093A1
• Time in Patent Office: 676 Days
• Field of Search: 705/5, 705/6, 705/10
• US Class Current: 705/7.31
• CPC Class Codes: G06Q 10/02 Reservations, e.g. for tickets, services or events
• CPC Class Codes: G06Q 10/025 Coordination of plural reservations, e.g. plural trip segments, transportation combined with accommodation
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities

Abstract

An airline origin-destination revenue management software system supports decisions to accept or deny request for booking on airline seats by comparing the fare for the request with a minimum acceptable fare predetermined by the system. In order to determine the minimum acceptable fare, the system typically solves an optimization problem that takes as input the airline'"'"'s flight network schedule, fares and forecasted passenger demands for all products available for booking on the airline'"'"'s flight network, and available capacity on each of the airline'"'"'s scheduled flights.

Claims

1. A method for determining an estimated origin-destination service product unconstrained demand, comprising the steps of:

retrieving a segment level unconstrained demand forecast for each flight segment within an origin-destination pair, wherein the flight segment comprises a segment having a single take-off and a single landing for an aircraft;

generating with a computer processor a consumer preference for origin-destination service products;

generating with the computer processor a network flight schedule by determining which available service products use one of the flight segments based on the segment level unconstrained demand forecast and the consumer preference for origin-destination service products, wherein the network flight schedule comprises a determination whether a service product uses a particular flight segment;

retrieving a historical origin-destination pair demand;

receiving a historical demand adjustment factor, wherein the historical demand adjustment factor is greater than 1 if a historical demand level is given greater weight than the segment level unconstrained demand forecast;

determining an estimated origin-destination unconstrained demand based on the segment level unconstrained demand forecast, the historical origin-destination pair demand, the network flight schedule, the consumer preference for origin-destination service products within the origin-destination pair, and the historical demand adjustment factor; and

computing the estimated origin-destination service product unconstrained demand based on the estimated origin-destination unconstrained demand and the consumer preference for origin-destination service products.

2. The method of claim 1, wherein the historical demand adjustment factor is equal to 0.1.

3. The method of claim 1, where the historical demand adjustment factor is less than 1 and greater than 0 if a historical demand level is given less weight as compared to the segment level unconstrained demand forecast.

4. The method of claim 1, wherein the historical origin-destination pair demand comprises a scaled historical origin-destination pair demand.

5. The method of claim 4, further comprising the step of generating the scaled historical origin-destination pair demand by:

retrieving an origin city for the origin-destination pair;

retrieving a destination city for the origin-destination pair;

determining if the destination city is an important destination city by identifying a percentage of consumers who begin at the origin city and end at the destination city as compared to all consumers beginning at the origin city;

generating with the computer processor a deflator for the origin city by dividing a non-terminating passenger count by an originating passenger count;

generating an estimated total traffic originating in the origin city by dividing a total load by the deflator;

generating with the computer processor the scaled historical origin-destination pair demand for the important city by calculating the product of the estimated total traffic originating in the origin city and the percentage of the consumers who begin at the origin city and end at the destination city;

generating with the computer processor used demand by summing the scaled historical origin-destination pair demand for each of the important cities;

generating an uncounted destination cities by summing the number of unimportant cities; and

determining the scaled historical origin-destination pair demand for the unimportant cities by calculating the difference between estimated total traffic and used demand and dividing the difference by the uncounted destination cities.

6. The method of claim 5, wherein the destination city is an important destination city if the destination city is within the 80^th percentile of total booked passengers departing the origin city.

7. The method of claim 1, wherein the step of computing an estimated origin-destination service product unconstrained demand comprises the steps of:

determining the available service products within the origin-destination pair;

computing the percentage of consumers who have historically used each of the available service products; and

computing with the computer processor the estimated origin-destination service product unconstrained demand by calculating the product of the estimated origin-destination unconstrained demand and the percentage of consumers who have historically used each of the available service products.

8. The method of claim 1, wherein the step of generating the network flight schedule comprises the steps of:

retrieving one of the flight segments from a flight segment list;

retrieving a service product from a service product list;

determining with the computer processor if the flight segment is used by the service product;

placing a 1 in a position representing the flight segment and the service product in the network flight schedule matrix if the flight segment is used by the service product;

placing a 0 in the position representing the flight segment and the service product in the network flight schedule matrix if the flight segment is not used by the service product.

9. A method for determining an estimated origin-destination service product unconstrained demand (d), comprising the steps of:

retrieving a segment level unconstrained demand forecast (v);

retrieving a consumer preference for origin-destination service products (Q);

generating with a computer processor a network flight schedule (N) by determining if a flight segment is included in a service product;

retrieving a historical origin-destination pair demand (w′

);

determining with the computer processor an estimated origin-destination unconstrained demand (w) using the formula

10. The method of claim 9, wherein the determining step comprises the steps of:

inputting the historical demand adjustment factor (α

); and

generating with the computer processor the estimated origin-destination unconstrained demand (w), using the formula

11. The method of claim 10, wherein the historical demand adjustment factor (α

) is 0.1.

12. The method of claim 10, wherein the historical demand adjustment factor (α

) is greater than 1 if a historical demand level is given a greater weight than the segment level unconstrained demand forecast (v).

13. The method of claim 10, wherein the historical demand adjustment factor (α

) is less than 1 and greater than 0 if a historical demand level is given less weight as compared to the segment level unconstrained demand forecast (v).

14. The method of claim 9, wherein the historical origin-destination pair demand (w′

) comprises a scaled historical origin-destination pair demand (ŵ

), and determining the scaled historical origin-destination pair demand (ŵ

) comprises the steps of;

retrieving an origin city for the origin-destination pair;

retrieving a destination city for the origin-destination pair;

determining if the destination city is an important destination city by identifying a percentage of consumers who begin at the origin city and end at the destination city as compared to all consumers beginning at the origin city;

generating with the computer processor a deflator for the origin city by dividing a non-terminating passenger count by an originating passenger count;

generating with the computer processor an estimated total traffic originating in the origin city by dividing a total load by the deflator;

generating with the computer processor the scaled historical origin-destination pair demand (ŵ

) for the important city by calculating the product of the estimated total traffic originating in the origin city and the percentage of the consumers who begin at the origin city and end at the destination city;

generating with the computer processor used demand by summing the scaled historical origin-destination pair demand (ŵ

) for each of the important cities;

generating with the computer processor an uncounted destination cities by summing to number of unimportant cities; and

determining with the computer processor the scaled historical origin-destination pair demand (ŵ

) for the unimportant cities by calculating the difference between estimated total traffic and used demand and dividing the difference by the uncounted destination cities.

15. The method of claim 14, wherein the destination city is an important destination city if the destination city is within the 80^th percentile of total booked passengers departing the origin city.

16. The method of claim 9, wherein the step of computing an estimated origin-destination service product unconstrained demand (d) comprises the steps of:

determining the available service products within the origin-destination pair;

computing with the computer processor the percentage of consumers who have historically used each of the available service products; and

computing with the computer processor the estimated origin-destination service product unconstrained demand (d) by calculating the product of the estimated origin-destination unconstrained demand (w) and the percentage of consumers who have historically used each of the available service products.

17. The method of claim 9, wherein the step of generating the network flight schedule (N) comprises the steps of:

retrieving one of the flight segments from a flight segment list;

retrieving a service product from a service product list;

determining with the computer processor if the flight segment is used by the service product;

placing a 1 in the position representing the flight segment and the service product in the network flight schedule (N) matrix if the flight segment is used by the service product;

placing a 0 in the position representing the flight segment and the service product in the network flight schedule (N) matrix if the flight segment is not used by the service product.

18. A method for determining an estimated origin-destination service product unconstrained demand (d), comprising the steps of:

inputting a segment level unconstrained demand forecast (v);

inputting a consumer preference for products within an origin-destination pair (Q);

inputting a historical origin-destination pair demand (w′

);

generating with a computer processor a network flight schedule (N) by determining if a flight segment is included in a service product;

determining with the computer processor an estimated origin-destination unconstrained demand (w) using the formula

19. The method of claim 18, wherein the determining step comprises the steps of;

inputting a historical demand adjustment factor (α

); and

generating with the computer processor an estimated origin-destination unconstrained demand (w), using the formula

20. The method of claim 19, wherein the historical demand adjustment factor (α

) is 0.1.

21. The method of claim 19, wherein the historical demand adjustment factor (α

) is greater than 1 if a historical demand level is given a greater weight than the segment level unconstrained demand forecast (v).

22. The method of claim 19, wherein the historical demand adjustment factor (α

) is less than 1 and greater than 0 if a historical demand level is given less weight as compared to the segment level unconstrained demand forecast (v).

23. The method of claim 18, wherein the historical origin-destination pair demand (w′

) comprises a scaled historical origin-destination pair demand (ŵ

).

24. The method of claim 23, further comprising the step of generating scaled historical origin-destination pair demand (ŵ

) by;

retrieving an origin city for the origin-destination pair;

retrieving a destination city for the origin-destination pair;

determining if the destination city is an important destination city by identifying a percentage of consumers who begin at the origin city and end at the destination city compared to all consumers beginning at the origin city;

generating with the computer processor a deflator for the origin city by dividing a non-terminating passenger count by an originating passenger count;

generating with the computer processor an estimated total traffic originating in the origin city by dividing a total load by the deflator;

generating with the computer processor the scaled historical origin-destination pair demand (ŵ

) for the important city by calculating the product of the estimated total traffic originating in the origin city and the percentage of the consumers who begin at the origin city and end at the destination city;

generating used demand by summing the scaled historical origin-destination pair demands (ŵ

) for each of the important cities;

generating with the computer processor an uncounted destination cities by summing the number of unimportant cities; and

determining with the computer processor the scaled historical origin-destination pair demand (ŵ

) for the unimportant cities by taking the difference between the estimated total traffic and the used demand and dividing the difference by the uncounted destination cities.

25. The method of claim 24, wherein the destination city is an important destination city if the destination city is within the 80^th percentile of total booked passengers departing the origin city.

26. The method of claim 18, wherein the step of computing an estimated origin-destination service product unconstrained demand (d) comprises the steps of:

determining the available service products within the origin-destination pair;

computing the percentage of consumers who have historically used each of the available service products; and

computing with the computer processor the estimated origin-destination service product unconstrained demand (d) by calculating the product of to estimated origin-destination unconstrained demand (w) and percentage of consumers who have historically used each of the available service products.

27. The method of claim 18, wherein the network flight schedule (N) comprises the steps of:

retrieving one of the flight segments from a flight segment list;

retrieving a service product from a service product list;

determining with the computer processor if the flight segment is used by the service product;

placing a 1 in the position representing die flight segment and the service product in the network flight schedule (N) matrix if the flight segment is used by the service product;

placing a 0 in the position representing the flight segment and the service product in the network flight schedule (N) matrix if the flight segment is not used by the service product.

28. A computer-readable medium for determining an estimated origin-destination service product unconstrained demand, said medium having computer-executable instructions for performing the steps comprising:

receiving a segment level unconstrained demand forecast for each flight segment within an origin-destination pair, wherein the flight segment comprises a segment having a single take-off and a single landing for an aircraft;

generating a consumer preference for origin-destination service products within the origin-destination pair;

generating a network flight schedule, wherein the network flight schedule comprises a determination whether a service product uses a particular flight segment;

receiving a historical origin-destination pair demand;

receiving a historical demand adjustment factor, wherein the historical demand adjustment factor is greater than 1 if a historical demand level is given greater weight than the segment level unconstrained demand forecast;

determining an estimated origin-destination unconstrained demand based on the segment level unconstrained demand forecast, the historical origin-destination pair demand, the network flight schedule, the consumer preference for origin-destination service products within the origin-destination pair, and the historical demand adjustment factor; and

computing the estimated origin-destination service product unconstrained demand.

29. A computer-readable medium for determining an estimated origin-destination service product unconstrained demand, said medium having computer-executable instructions for performing the steps comprising:

receiving a segment level unconstrained demand forecast for each flight segment within an origin-destination pair, wherein the flight segment comprises a segment having a single take-off and a single landing for an aircraft;

generating a consumer preference for origin-destination service products within the origin-destination pair;

generating a network flight schedule, wherein the network flight schedule comprises a determination whether a service product uses a particular flight segment;

receiving a historical origin-destination pair demand;

receiving a historical demand adjustment factor, wherein the historical demand adjustment factor is less than 1 and greater than 0 if a historical demand level is given less weight as compared to the segment level unconstrained demand forecast;

determining an estimated origin-destination unconstrained demand based on the segment level unconstrained demand forecast, the historical origin-destination pair demand, the network flight schedule, the consumer preference for origin-destination service products within the origin-destination pair, and the historical demand adjustment factor; and

computing the estimated origin-destination service product unconstrained demand.