System and method for optimization process repeatability in an on-demand computing environment

• US 20060058991A1
• Filed: 09/16/2004
• Published: 03/16/2006
• Est. Priority Date: 09/16/2004
• Status: Active Grant

• Current Assignee: International Business Machines Corporation
• Original Assignee: International Business Machines Corporation
• Inventors: Xi, Haifeng, Kumar, Tarun, Parija, Gyana R.
• Granted Patent: US 8,036,921 B2
• Time in Patent Office: Days
• US Class Current: 703/22
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 10/063 Operations research, analysis or management

Abstract

A method (and system) of providing optimization repeatability in an on-demand computing environment removes variability in an optimization model instance and can be exemplarily implemented in a service architecture. The method and system receives a plurality of physical data instances, which are different representations of the same logical data model, and transforms the plurality of physical data instances into a normalized physical data instance, which can be combined with an optimization model to form a unique optimization model instance, thereby providing repeatability in solving optimization problems.

Claims

1. A method of providing optimization repeatability in an on-demand computing environment, comprising:

removing variability in an optimization model instance.

2. The method according to claim 1, wherein said removing variability in said optimization model instance, comprises:

receiving a plurality of physical data instances; and

generating a normalized physical data instance from said plurality of physical data instances.

3. The method according to claim 2, wherein said normalized physical data instance includes a unique ordering of data elements.

4. The method according to claim 3, wherein at least one data element of said unique ordering of data elements includes a collection of objects, and

wherein said collection of objects is sorted by a predetermined object attribute.

5. The method according to claim 3, wherein said unique ordering of data elements comprises hierarchical ordering.

6. The method according to claim 5, wherein said hierarchical ordering comprises a TreeMap structure.

7. The method according to claim 2, further comprising:

determining said optimization model instance, based on an optimization model and said normalized physical data instance.

8. The method according to claim 7, further comprising:

providing a plurality of solvers; and

selecting one of said plurality of solvers based on said optimization model instance.

9. The method according to claim 1, wherein said removing variability in said optimization model instance, comprises:

generating a unique normalized physical data instance from a plurality of physical data instances.

10. The method according to claim 9, wherein said removing variability in said optimization model instance further comprises:

determining said optimization model instance by combining said single unique normalized physical data instance with an optimization model.

11. The method according to claim 1, wherein said optimization model instance comprises a reproduced optimized model instance.

12. The method according to claim 1, wherein said optimization model instance comprises a unique optimized model instance.

13. The method according to claim 1, wherein said optimization model instance comprises a unique MPS (mathematical programming system) format input file.

36. The method according to claim 9, wherein said unique normalized physical data instance comprises a single normalized physical data instance.

14. An optimization method for an on-demand computing environment, comprising:

receiving a plurality of physical data instances;

transforming said plurality of physical data instances into a normalized physical data instance; and

obtaining an optimization model instance based on an optimization model and said normalized physical data instance.

15. A system for providing optimization repeatability in an on-demand computing environment, comprising:

a data transformer that removes variability in an optimization model instance.

16. The system according to claim 15, wherein said data transformer comprises:

a receiver that receives a plurality of physical data instances; and

a generator that generates a normalized physical data instance from said plurality of physical data instances.

17. The system according to claim 16, further comprising:

a determiner that receives said normalized physical data instance from said data transformer and determines said optimization model instance, based on an optimization model and said normalized physical data instance.

18. The system according to claim 16, wherein said normalized physical data instance includes a unique ordering of data elements.

19. The system according to claim 18, wherein at least one data element of said unique ordering of data elements includes a collection of objects, and

wherein said collection of objects is sorted by a predetermined object attribute.

20. The system according to claim 18, wherein said unique ordering of data elements comprises hierarchical ordering.

21. The system according to claim 20, wherein said hierarchical ordering comprises a TreeMap structure.

22. The system according to claim 15, wherein said data transformer removes variability in said optimization model instance and generates a unique normalized physical data instance from a plurality of physical data instances.

23. The system according to claim 22, further comprising:

a determiner that determines said optimization model instance based on said single unique normalized physical data instance and an optimization model.

24. The system according to claim 15, wherein said optimization model instance comprises a reproduced optimized model instance.

25. The system according to claim 15, wherein said optimization model instance comprises a unique optimized model instance.

26. The system according to claim 15, wherein said optimization model instance comprises a unique mathematical programming system format input file.

37. The method according to claim 22, wherein said unique normalized physical data instance comprises a single normalized physical data instance.

27. A system for providing optimization repeatability in an on-demand computing environment, comprising:

a receiver that receives a plurality of physical data instances;

a transformer that transforms said plurality of physical data instances into a normalized physical data instance; and

a determiner that determines an optimized model instance based on an optimization model and said normalized physical data instance.

28. The system according to claim 27, further comprising:

a solver that receives said optimization model instance from said determiner and determines optimization based on said optimization model instance.

29. The system according to claim 27, further comprising:

a plurality of solvers,

wherein one of said plurality of solvers is selectable for receiving said optimization model instance from said determiner and generating an optimal solution based on said optimization model instance.

30. A system for providing optimization repeatability in an on-demand computing environment, comprising:

means for receiving a plurality of physical data instances and generating a normalized physical data instance from said plurality of physical data instances;

means for determining a unique optimization model instance, based on said normalized physical data instance and an optimization model; and

means for receiving said unique optimization model instance and generating an optimal solution based on said unique optimization model instance.

31. A signal-bearing medium tangibly embodying a program of machine-readable instructions executable by a digital processing apparatus to perform a method of providing optimization process repeatability in an on-demand computing environment, the method comprising:

removing variability in an optimization model instance.

32. A method of deploying computing infrastructure in which computer-readable code is integrated into a computing system, and combines with said computing system to perform a method of providing optimization process repeatability in an on-demand computing environment, said method comprising:

removing variability in an optimization model instance.

33. The method according to claim 32, wherein said removing variability in said optimization model instance, comprises:

providing a plurality of services including at least one of an optimization service, a data transformation service, and a solver service.

34. The method according to claim 33, further comprising:

simultaneously deploying to a plurality of users of said computing system a change of service of at least one of said plurality of services.

35. The method according to claim 34, wherein said change of service includes at least one of upgrading, removing, altering, and replacing at least one of said plurality of services.