Design optimization system and method
First Claim
Patent Images
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.
1 Assignment
0 Petitions
Accused Products
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.
-
Citations
46 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. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17)
-
-
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. - View Dependent Claims (19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33)
-
-
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 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. - View Dependent Claims (35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46)
-
-
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:
-
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.
Specification