Common samplings created in a constrained design space might produce infeasible sample points if the underlying formulation disregards the constraints. DesignXplorer provides a sampling type that takes into account the constraints defined for input parameters.
DesignXplorer generates sampling:
Subject to:
The symbols and
denote the vectors of the continuous and the integer variables,
respectively. DesignXplorer allows you to define a constrained design space with linear or
non-linear constraints.
Examples of linear constraints:
Examples of non-linear constraints:
The constraint sampling is a heuristic method based on Shifted-Hammersley (Screening) and MISQP sampling methods.
For a given screening of sample points generated in the hypercube of input parameters, only a
part of this sampling (
) is within the constrained design space (feasible domain).
To obtain a constraint sampling that is close to a uniform sampling of points, DesignXplorer solves this problem first:
If the sampling is completed.
If , only the
furthermost sample points are kept.
Otherwise, DesignXplorer needs to create additional points to reach the sample points.
The Screening is not guaranteed to find either enough sample points or at least one feasible point. For this reason, DesignXplorer solves an MISQP problem for each constraint on input parameters, as follows:
Subject to:
In the best case, you obtain feasible points.
If is the center of mass of the feasible points, then:
DesignXplorer then solves the new MISQP problem:
Subject to:
Once the point closest to the center mass of the feasible points has been found, DesignXplorer projects part of the infeasible points onto the skin of the feasible domain.
Given an infeasible point and feasible point
, you can build a new feasible point
:
To find close to the skin of the feasible domain, the algorithm starts with
equals 1, and decreases
value until
is feasible.

To optimize the distribution of point on the skin of the feasible domain, the algorithm chooses the furthermost infeasible points in terms of angles:
If is the first infeasible point processed, the next one is:
Once enough points have been generated on the skin of the feasible domain, the algorithm generates internal points. The internal points are obtained by combination of internal points and points generated on the skin of the feasible domain.