Sensitivity analysis quantifies the importance of the design variables with respect to the model responses. This approach is based on a design exploration. For this purpose classical Design of Experiments schemes as Full factorial, Central composite and D-optimal designs and random sampling methods are available. The latter are Monte Carlo Simulation which is pure random sampling and Latin Hypercube Sampling, which helps to explore the design space with a smaller number of samples and reduces spurious dependencies between the design variables. Further available sampling algorithms are Koshal (linear, quadratic), Full combinatorial, Box-Behnken, Star points, Advanced Latin Hypercube Sampling, Space filling Latin Hypercube Sampling and Sobol Sequence. By applying random sampling procedures the exploration assumes uniformly distributed and independent design variables. The exploration designs are evaluated by the solver and for each design the corresponding response values are obtained. Based on this data the sensitivity measures as Coefficients of Correlation and polynomial based Coefficients of Importance are estimated. Failed designs are neglected in this statistical evaluation. Based on the Coefficient of Prognosis as quality measure the Metamodel of Optimal Prognosis approach detects for each specified solver response the optimal approximation model using the optimal subspace of important variables. MOP-based variable sensitivities are quantified by variance based sensitivity indices.
Further information about methods of sensitivity analysis used in optiSLang can be found here.
Relevant Parameters
Only parameters with deterministic properties (Optimization or Opt.+Stoch. parameter) are relevant for the sensitivity analysis.
Initialization Options
To access the options shown in the following table, double-click the Sensitivity system on the Scenery pane and switch to the Dynamic sampling tab.
Option | Description |
---|---|
Dynamic sampling | This check box is selected by default. When cleared, only start designs are evaluated and considered for the sensitivity analysis. |
Sampling type |
Select one of the following sampling methods: |
Number of samples | Displays the number of samples or number of levels used by the selected sampling type. The number of levels is displayed for methods where the number of calculations depend on the number of inputs. |
Additional Options
To access the options shown in the following table, in any tab, click
.Option | Description |
---|---|
Limit maximum in parallel |
Controls the resource usage of nodes in the system. When the check box is cleared (default), a value is chosen to ensure the best possible utilization of the child nodes. When the check box is selected, set the value manually to specify how many designs are sent to child nodes, limiting the maximum degree of parallelism for all children. Ansys recommends keeping the check box clear. |
Auto-save behavior |
Select one of the following options:
The project, including the database, is auto-saved (depending on defined interval) after calculating this node/system (either when the calculation succeeds or fails). By default, all parametric and algorithm systems have selected, all other nodes have selected. |