Design optimization system and method

• US 7,756,688 B2
• Filed: 05/09/2005
• Issued: 07/13/2010
• Est. Priority Date: 05/10/2004
• Status: Active Grant

• Current Assignee: The Board of Trustees of Michigan State University (Michigan State University)
• Original Assignee: The Board of Trustees of Michigan State University (Michigan State University)
• Inventors: Halepatali, Praveen, Sidhu, Ran Mohan Jit Singh, Averill, Ronald C.
• Primary Examiner(s): Craig; Dwin M
• Application Number:
  US11/579,711
• Publication Number:
  US 20070244678A1
• Time in Patent Office: 1,891 Days
• Field of Search: 703/1, 703/2, 700/97, 700/89, 706/13
• US Class Current: 703/2
• CPC Class Codes: G06F 2111/06 Multi-objective optimisation, e.g. Pareto optimisation using simulated annealing [SA], ant colony algorithms or genetic algorithms [GA]
• CPC Class Codes: G06F 2111/08 Probabilistic or stochastic CAD
• CPC Class Codes: G06F 30/00 Computer-aided design [CAD]
• CPC Class Codes: G06F 8/20 Software design
• 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

Abstract

A design optimization system includes an initial design evaluation module evaluating a system model using initial subsystem designs to extract interactions between subsystem models and/or between the system and the subsystem models. An updating module updates the interactions for the subsystem models, and an optimization module performs subsystem design optimization using the subsystem model and most recently updated interactions, thereby obtaining an updated subsystem design.

Claims

1. A method of manufacturing a mechanical part, comprising:

(a) constructing a system model for a mechanical system having the mechanical part and at least one other mechanical part of the mechanical system, the system model comprised of multiple subsystem models including a given subsystem model that includes the mechanical part;

(b) evaluating the system model using initial subsystem designs and extracting interactions between the given subsystem model and other subsystems models;

(c) updating the interactions for the given subsystem model;

(d) performing subsystem design optimization on the given subsystem model using the subsystem model and most recently updated interactions while varying magnitude of stochastic variability on the interactions in a range on the order of 5% to 20% during the subsystem design optimization, thereby obtaining an updated subsystem design; and

(e) manufacturing the mechanical part based on the updated subsystem design.

2. The method of claim 1, wherein step (c) includes updating the interactions for the subsystem model i at iteration k according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}$

where û

_s(x,y,z,t) are the interactions on the subsystem boundary Γ

_subsystem(i) at iteration s, and {tilde over (w)}_s(x,y,z,t) are weight functions.

3. The method of claim 2, wherein said interactions are primary variables relating to generalized displacements.

4. The method of claim 2, wherein said interactions are secondary variables relating to generalized forces.

5. The method of claim 2, wherein said interactions are primary variables relating to generalized displacements and secondary variables relating to generalized forces.

6. The method of claim 2, wherein the weight functions have predetermined spatial and temporal distributions.

7. The method of claim 2, wherein the weight functions have magnitudes that are varied stochastically within a selected range during performance of the subsystem design optimization.

8. The method of claim 1, further comprising:

(f) evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models; and

(g) evaluating and selectively obeying a termination condition.

9. The method of claim 8, further comprising selectively returning to step (c) and updating the interactions if a solution is improving according to an objective function, thereafter proceeding to steps (d), then (e), and then (f).

10. The method of claim 8, further comprising selectively returning to step (c) if the solution is not improving according to the objective function, thereafter proceeding to steps (d), then (e), and then (f), wherein at least one of additional search at higher variation is performed, or the interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}.$

11. The method of claim 1, further comprising selectively returning to step (d) if a solution is not improving according to an objective function.

12. The method of claim 11, wherein additional search at higher variation is performed.

13. The method of claim 11, wherein interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}.$

14. The method of claim 1, further comprising evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models.

15. The method of claim 1, further comprising evaluating and selectively obeying a termination condition.

16. The method of claim 1, further comprising selectively returning to step (c) and updating the interactions if a solution is improving according to an objective function.

17. The method of claim 1, wherein the mechanical part is a motor vehicle part, and the mechanical system is a motor vehicle or part thereof having multiple, interacting mechanical parts including the mechanical part.

18. A computer-implemented design optimization method, comprising:

(a) the computer evaluating a system model comprised of multiple subsystem models using initial subsystem designs and extracting interactions between a given subsystem model and other subsystem models;

(b) the computer updating the interactions for the given subsystem model; and

(c) the computer performing subsystem design optimization on the given subsystem model using the subsystem model and most recently updated interactions while limiting magnitude of stochastic variability on the interactions in a during the subsystem design optimization within a selected range having more than one value, thereby obtaining an updated subsystem design.

19. The method of claim 18, wherein step (b) includes updating the interactions for the subsystem model i at iteration k according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}$

where û

_s(x,y,z,t) are the interactions on the subsystem boundary Γ

_subsystem(i) at iteration s, and {tilde over (w)}_s(x,y,z,t) are weight functions.

20. The method of claim 19, wherein said interactions are primary variables relating to generalized displacements.

21. The method of claim 19, wherein said interactions are secondary variables relating to generalized forces.

22. The method of claim 19, wherein said interactions are primary variables relating to generalized displacements and secondary variables relating to generalized forces.

23. The method of claim 19, wherein the weight functions have predetermined spatial and temporal distributions.

24. The method of claim 19, wherein the weight functions have magnitudes that are varied stochastically within a selected range during the optimization process.

25. The method of claim 18, further comprising:

(d) evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models;

(e) evaluating and selectively obeying a termination condition.

26. The method of claim 25, further comprising selectively returning to step (b) and updating the interactions if a solution is improving according to an objective function, thereafter proceeding to steps (c), then (d), and then (e).

27. The method of claim 25, further comprising selectively returning to step (c) if the solution is not improving according to the objective function, thereafter proceeding to steps (c), then (d), and then (e), wherein at least one of additional search at higher variation is performed, or the interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}$

28. The method of claim 18, further comprising selectively returning to step (c) if a solution is not improving according to an objective function.

29. The method of claim 28, wherein additional search at higher variation is performed.

30. The method of claim 28, wherein interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}$

31. The method of claim 18, further comprising evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models.

32. The method of claim 18, further comprising evaluating and selectively obeying a termination condition.

33. The method of claim 18, further comprising selectively returning to step (b) and updating the interactions if a solution is improving according to an objective function.

34. A computer-implemented design optimization method, comprising:

(a) the computer evaluating a system model comprised of multiple subsystem models using initial subsystem designs and extracting interactions between a given subsystem model and other subsystem models;

(b) the computer updating the interactions for the given subsystem model i at iteration k according to

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}$

where û

_s(x,y,z,t) are the interactions on the subsystem boundary Γ

_subsystem(i) at iteration s, and {tilde over (w)}_s(x,y,z,t) are weight functions; and

(c) the computer performing subsystem design optimization on the given subsystem model using the subsystem model and most recently updated interactions while the magnitudes of the weight functions have predetermined spatial and temporal distributions and are varied stochastically during the subsystem design optimization in a range having more than one value, thereby obtaining an updated subsystem design.

35. The method of claim 34, wherein said interactions are primary variables relating to generalized displacements.

36. The method of claim 34, wherein said interactions are secondary variables relating to generalized forces.

37. The method of claim 34, wherein said interactions are primary variables relating to generalized displacements and secondary variables relating to generalized forces.

38. The method of claim 34, further comprising:

(d) evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models;

(e) evaluating and selectively obeying a termination condition.

39. The method of claim 38, further comprising selectively returning to step (b) and updating the interactions if a solution is improving according to an objective function, thereafter proceeding to steps (c), then (d), and then (e).

40. The method of claim 38, further comprising selectively returning to step (c) if the solution is not improving according to the objective function, thereafter proceeding to steps (c), then (d), and then (e), wherein at least one of additional search at higher variation is performed, or the interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}.$

41. The method of claim 34, further comprising selectively returning to step (c) if a solution is not improving according to an objective function.

42. The method of claim 41, wherein additional search at higher variation is performed.

43. The method of claim 41, wherein interactions are scaled back toward those of previous iterations according to:

${\hat{u}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}=\underset{s=1}{\overset{k}{\sum <br><br>}}<br><br>{\tilde{w}}_s<br><br>{\hat{u}}_s<br><br><br><br>\mathrm{on}<br><br><br><br>{\mathrm{\Gamma <br><br>}}_{\mathrm{subsystem}<br><br><br><br>\left(i\right)}.$

44. The method of claim 34, further comprising evaluating one or more system models using one or more updated subsystem designs to extract new interactions for one or more subsystem models.

45. The method of claim 34, further comprising evaluating and selectively obeying a termination condition.

46. The method of claim 34, further comprising selectively returning to step (b) and updating the interactions if a solution is improving according to an objective function.