DESOBJ
Subcase Information Entry To select a single response definition as the objective function of an optimization, or to select system response definitions when the objective function is the least squares sum of these definitions. The DESOBJ command also indicates if this response is to be minimized or maximized.
Format
DESOBJ (type) = integer, PROB
Definitions
Argument | Options | Description |
---|---|---|
type | <MIN, MAX, MINP, MAXP> | |
integer | <ID> No default |
|
PROB | <Probability> <50.0 < Real < 100.00> |
|
Comments
- Only one DESOBJ card can be present.
- For global DRESP1 responses (responses which are not subcase dependent) or DRESP2 or DRESP3 responses containing either DRESP1L/DRESP2L data or global DRESP1 responses, the DESOBJ data must be above the first SUBCASE statement.
- If the DESOBJ data references responses that are subcase specific, the DESOBJ statement must be within the appropriate subcase definition and the subcase must be of the appropriate type.
- If the DESOBJ entry is referenced
by a DESOBJ Subcase entry, a least squares objective function
is used in the optimization. The objective function is the sum of the squared,
weighted (
), normalized differences between the target
responses (
)and those calculated by the finite element
analysis (
).
(1) - DESOBJ entries must have unique identification numbers with respect to DRESP1, DRESP2, and DRESP3 entries.
- DRESP1, DRESP2, and DRESP3 entries referenced by the DSYSID entry can define only a single response per subcase when the DESOBJ formulation is used.
- Time dependent responses should not be referenced by the DESOBJ entry. The minmax formulation should be used for optimization problems that have time-dependent responses as the objective functions. The minmax formulation can be selected using the MINMAX or MAXMIN Subcase Information Entry.
- The MAXP, MINP, and PROB options can be input during a Reliability-based Design Optimization run. MINP and MAXP are not supported if random design variables or random parameters are not defined in the model.
- This entry is represented as an optimization objective in HyperMesh.