PCA-based sensitivity analysis of response fields using LS-OPT®

When performing an optimization, it is important to avoid introducing unnecessary variables that do not impact the design objectives and constraints. Such variables increase the design space size and lead to unnecessary sample evaluations, which can significantly increase the overall computation time or cost. A sensitivity analysis can be performed to quantify the significance of the variables; only the important variables are then used in the sampling and optimization, thus reducing the computational cost.