Our website uses cookies. By using the website you agree ot its use. More information can be found in our privacy policy.

Response Surface Approximate Optimization (RSA)

Response Surface Approximate Optimization (RSA)

The RSA component combines Optimization (OPT), Design of Experiments (DOE), and Response Surface Modeling (RSM) to improve optimization efficiency. The approximation is used as a surrogate for the underlying computationally expensive analyses and is incrementally refined as the optimization proceeds. The user can choose the DOE technique, the optimization algorithm, and the approximation model to use with the RSA component.

An Approximate Surface

The following approximation models are available:

  • Linear
  • Mixed: Linear + Interaction
  • Mixed: Linear + Quadratic
  • Full Quadratic
  • Forward Stepwise Regression