When using Sparse Grid auto-refinement capabilities, there are four main steps to the process:
Setting up the DOE
Setting up the response surface
Updating the response surface
Generating verification points
To set up the DOE for Sparse Grid auto-refinement:
In the Outline view for the Design of Experiments cell, select the node.
Under Design of Experiments in the Properties view, set Design of Experiments Type to .
Either preview or update the DOE to validate your selections. This causes the Response Surface Type property for Response Surface cells in downstream design exploration systems to default to .
Note: If the Design of Experiments cell is shared by multiple systems, the DOE definition applies to the Response Surface cells for each of those systems.
To set up refinement properties for a Sparse Grid auto-refinement:
In the Outline view for the Response Surface cell, select the Response Surface node.
Under Meta Model in the Properties view, verify that Response Surface Type is set to .
Under Refinement in the Properties view:
For Maximum Relative Error (%), enter the value to be allowed for all of the output parameters. The smaller the value, the more accurate the response surface.
For Maximum Depth, enter the number of grid levels that can be created in a given direction. If needed, you can adjust this value later, according to your update results.
For Maximum Number of Refinement Points, enter the maximum number of refinement points that can be generated as part of the refinement process.
For more information, see Sparse Grid Auto-Refinement Properties.
Once the DOE and response surface are defined, run the operation on the response surface to adaptively generate the Sparse Grid response surface.
At any time, you can click the red stop button in the Progress view to interrupt the Sparse Grid refinement so that you can change properties or see partial results. The refinement points that have already been calculated are visible, and the displayed charts are based on the response surface's current level of refinement. The refinement points that have not been calculated display the Update Required icon to indicate which output parameters must be updated.
If a design point fails during the refinement process, Sparse Grid refinement stops in the area where the failed design point is located, but refinement continues in the rest of the parameter space to the degree possible. Failed refinement points are indicated by the Update Failed, Update Required icon. You can attempt to pass the failed design points by updating the response surface once again.
The auto-refinement process continues until the Maximum Relative Error (%) objective is attained, the Maximum Depth limit is reached for all input parameters, the Maximum Number of Refinement Points is reached, or the response surface converges. If all output parameters have a Maximum Relative Error (%) that is higher than the Current Relative Error defined for them in the Properties view, the refinement is converged.
Note: If Sparse Grid refinement does not appear to be converging, you can do the following to accept the current level of convergence:
Stop the process.
Either set the value for Maximum Relative Error (%) slightly above the value for Current Relative Error (%) or set the value for Maximum Number of Refinement Points equal to the value for Number of Refinement Points.
Update the response surface again.
Like Kriging, Sparse Grid is an interpolation. To assess the quality of the response surface, verification points are needed. Once the Sparse Grid auto-refinement has converged, you should generate verification points to validate the response surface accuracy.
Note: Because Sparse Grid uses an auto-refinement process, you cannot add refinement points manually as with other DOE types. To generate verification points automatically, under Verification Points in the Properties view for the Response Surface cell, select the Generate Verification Points check box. Then, update the response surface.
For more information on assessing the quality of a response surface, see Goodness of Fit and Verification Points.