Design optimization

• US 20070255543A1
• Filed: 06/12/2007
• Published: 11/01/2007
• Est. Priority Date: 08/31/2001
• Status: Abandoned Application

• Current Assignee: Optimum Power Technology LP
• Original Assignee: Optimum Power Technology LP
• Inventors: Crandall, John, Chatfield, Glen
• Application Number:
  US11/811,673
• Publication Number:
  US 20070255543A1
• Time in Patent Office: Days
• US Class Current: 703/6
• CPC Class Codes: F02D 2041/1423 Identification of model or controller parameters
• CPC Class Codes: F02D 2041/1433 using a model or simulation of the system
• CPC Class Codes: F02D 41/1406 with use of a optimisation method, e.g. iteration
• CPC Class Codes: G05B 13/042 in which a parameter or coefficient is automatically adjusted to optimise the performance
• CPC Class Codes: G05B 17/02 electric
• 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 30/00 Computer-aided design [CAD]
• 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

An apparatus, system, and method for optimization. The apparatus, system, and method include a mechanism for ending an optimization when all models within one design tolerance of an optimum model have been simulated.

Claims

1. A computer readable medium containing instructions which, when executed by a processor, cause the processor to:

a. specify a model for performance of a device that includes an input variable having a tolerance and a step having an initial magnitude greater than the tolerance, and an output characteristic having an objective;

b. simulate the model with the input variable set to a selected value to arrive at a resulting output characteristic;

c. simulate the model with the input variable set to the selected value changed by the step to arrive at a second resulting output characteristic;

d. select the input variable value that results in an output characteristic that more nearly conforms to the objective;

e. reduce the step if the selected input variable value is unchanged at d; and

f. repeat c through e until the step reaches the tolerance.

2. The computer readable medium of claim 1, wherein the step reaches the tolerance when the step first becomes less than the tolerance.

3. The computer readable medium of claim 1, wherein the step reaches the tolerance when the step is equal to the tolerance.

4. The computer readable medium of claim 1, wherein the next reduction would cause the step to be less than the tolerance.

5. The computer readable medium of claim 1, wherein the next reduction would cause the step to be equal to the tolerance.

6. The computer readable medium of claim 1, further comprising:

running the model at the input variable value that results in an output characteristic that more nearly conforms to the objective plus one tolerance prior to terminating the method if the model was not previously run at the input variable value that results in an output characteristic that more nearly conforms to the objective plus one tolerance;

running the model at the input variable value that results in an output characteristic that more nearly conforms to the objective minus one tolerance prior to terminating the method if the model was not previously run at the input variable value that results in an output characteristic that more nearly conforms to the objective minus one tolerance; and

selecting one of the input variable value, the input variable value plus one tolerance, and the input variable value minus one tolerance that results in an output characteristic that more nearly conforms to the objective.

7. The computer readable medium of claim 1, further comprising not changing the step if the selected input variable value is changed at d.

8. The computer readable medium of claim 1, wherein the output characteristic that more nearly conforms to the objective is the output characteristic that most closely matches a predefined value of the objective.

9. The computer readable medium of claim 1, wherein the output characteristic that more nearly conforms to the objective is the output characteristic having the smallest value.

10. The computer readable medium of claim 1, wherein changed by a step includes subtracting the step from the selected value.

11. The computer readable medium of claim 1, wherein changed by a step includes adding the step to the selected value.

12. The computer readable medium of claim 1, wherein running the model with the input variable set to the selected value changed by the step includes running the model with the input variable set to the selected value plus the step and running the model with the input variable set to the selected value minus the step.

13. The computer readable medium of claim 1, further comprising setting a minimum boundary for the variable below which optimization will not be performed.

14. The computer readable medium of claim 1, further comprising setting a maximum boundary for the variable above which optimization will not be performed.

15. The computer readable medium of claim 1, wherein the model is a model of at least one of an apparatus, an apparatus related process, and a chemical process.

16. The computer readable medium of claim 1, further comprising specifying at least a second input variable in the model, the second input variable having a second tolerance and a second step and a second initial magnitude greater than the second tolerance.

17. A processor containing instructions that, when executed by the processor, cause the processor to:

a. specify a model for performance of a device that includes an input variable having a tolerance, a step having an initial magnitude greater than the tolerance, and an output characteristic having an objective;

b. simulating the model with the input variable set to a selected value to arrive at a resulting output characteristic;

c. simulating the model with the input variable set to the selected value changed by the step to arrive at a second resulting output characteristic;

d. select the input variable value that results in an output characteristic that more nearly conforms to the objective;

e. reduce the step if the selected input variable value is unchanged at d; and

f. repeat c through e until the step reaches the tolerance.

18. The processor of claim 17, further comprising a storage device.

19. The processor of claim 17, wherein at least one of the variable value that results in the chosen output characteristic value and the chosen output characteristic is stored in the storage device.

20. The processor of claim 17, wherein the storage device includes memory.

21. The processor of claim 17, wherein the output characteristic that more nearly conforms to the objective is the output characteristic that most closely matches a predefined value of the objective.

22. The processor of claim 17, wherein the model is a model of at least one of an apparatus, an apparatus related process, and a chemical process.

23. The processor of claim 17, further comprising specifying at least a second input variable in the model, the second input variable having a second tolerance and a second step and a second initial magnitude greater than the second tolerance.

24. The processor of claim 17, wherein the input variable is a dimension of a component of a machine and the model simulates the machine including that component.

25. A computer readable medium containing instructions that, when executed by a processor, cause the processor to:

a. specify a base model for performance of a device that includes a variable having a value;

b. specify a goal that identifies a characteristic of each model for performance of a device that is optimized, the best model being the model that produces the best result for the characteristic;

c. specify a tolerance that is a minimum amount that the variable can be changed;

d. specify a delta that is initially set to a maximum amount that the variable can be changed;

e. simulate the base model to obtain a result for the base model;

f. identify the current best model to be the base model;

g. create a plus model by setting the variable value to the current best model value plus the delta;

h. simulate the plus model to obtain a result for the plus model;

i. create a minus model by setting the variable value to the current best model value minus the delta;

j. simulate the minus model to obtain a result for the minus model;

k. setting a previous best model to the current best model and setting the current best model to one of the current best model, the plus model and the minus model having the best result for the characteristic;

l. repeating steps (g) through (l) if the current best model is different than the previous best model;

m. reducing the delta;

n. repeating steps (g) through (n) if the delta is greater than or equal to the tolerance; and

o. identifying the current best model as the optimum model.

26. A computer readable medium containing instructions that, when executed by a processor, cause the processor to:

a. specify a base model that includes two variables, a first variable having a first value and a second variable having a second value;

b. specifying a goal which identifies a characteristic of the base model that is to be optimized;

c. specifying a first tolerance that is a minimum amount that the first variable can be changed and a second tolerance that is a minimum amount that the second variable can be changed;

d. specifying a first delta that is initially set to a maximum amount that the first variable can be changed and a second delta that is initially set to a maximum amount that the second variable can be changed;

e. running the base model;

f. identifying the current best model to be the base model;

g. creating a first plus model by setting the first variable value to the current best model first value plus the first delta;

h. running the first plus model;

i. creating a first minus model by setting the first variable value to the current best model first value minus the first delta;

j. running the first minus model;

k. creating a second plus model by setting the second variable value to the current best model second value plus the second delta;

l. running the second plus model;

m. creating a second minus model by setting the second variable value to the current best model second value minus the second delta;

n. running the second minus model;

o. setting a previous best model to the current best model and setting the current best model to the best of the current best model, the first plus model, the first minus model, the second plus model and the second minus model;

p. repeating steps (g) through (p) if the current best model is different than the previous best model;

q. reducing the first delta and the second delta;

r. repeating steps (g) through (r) if the first delta is greater than or equal to the first tolerance or if the second delta is greater than or equal to the second tolerance; and

s. identifying the current best model as the optimum model.

27. A computer readable medium containing instructions that, when executed by a processor, cause the processor to:

a. specify a base model that includes two variables, a first variable having a first value and a second variable having a second value;

b. specify a goal which identifies a characteristic of the base model that is to be optimized;

c. specify a first tolerance that is a minimum amount that the first variable can be changed and a second tolerance that is a minimum amount that the second variable can be changed;

d. specify a first delta that is initially set to a maximum amount that the first variable can be changed and a second delta that is initially set to a maximum amount that the second variable can be changed;

e. running the base model;

f. identifying the current best model to be the base model;

g. creating a first plus model by setting the first variable value to the current best model first value plus the first delta;

h. running the first plus model;

i. creating a first minus model by setting the first variable value to the current best model first value minus the first delta;

j. running the first minus model;

k. creating a second plus model by setting the second variable value to the current best model second value plus the second delta;

l. running the second plus model;

m. creating a second minus model by setting the second variable value to the current best model second value minus the second delta;

n. running the second minus model;

o. creating a plus-minus model by setting the first variable value to the current best model first value plus the first delta and setting the second variable value to the current best model second value minus the second delta;

p. running the plus-minus model;

q. creating a minus-plus model by setting the first variable value to the current best model first value minus the first and setting the second variable value to the current best model second value plus the second delta;

r. running the minus-plus model;

s. creating a plus-plus model by setting the first variable value to the current best model first value plus the first delta and setting the second variable value to the current best model second value plus the second delta;

t. running the plus-plus model;

u. creating a minus-minus model by setting the first variable value to the current best model first value minus the first and setting the second variable value to the current best model second value minus the second delta;

v. running the minus-minus model;

w. setting a previous best model to the current best model and setting the current best model to the best of the current best model, the first plus model, the first minus model, the second plus model, the second minus model, the plus-minus model, the minus-plus model, the plus-plus model and the minus-minus model;

x. repeating steps (g) through (w) if the current best model is different than the previous best model;

y. reducing the first delta and the second delta;

z. repeating steps (g) through (z) if the first delta is greater than or equal to the first tolerance or if the second delta is greater than or equal to the second tolerance; and

aa. identifying the current best model as the optimum model.