Active Manifold and Model Order Reduction to Accelerate Multidisciplinary Analysis and Optimization - presented by Prof. Charbel Farhat

Active Manifold and Model Order Reduction to Accelerate Multidisciplinary Analysis and Optimization

Prof. Charbel Farhat

Prof. Charbel Farhat
Slide at 44:19
Charbel Sampling the Parameter Domain for Training: Greedy Procedure Generic procedure for training and building a global ROB V Initial point Random candidates step 0 : : randomly sample one point and construct initial global ROB at this point step j : j = 2, 3, ... Nsmp randomly pick N° candidate points using, for example, LHS at each candidate point, assess the accuracy of the global PROM associated with the current global ROB using a residual-based error indicator
Share slide
Summary (AI generated)

If a point has the highest residual, it means that the current model is not accurate at that point. To improve accuracy, that point should be added to the sample points. Retrain on the snapshots from these two points and repeat this process until a non-uniform sampling is achieved. This adaptive method considers feedback and helps reduce the curse of dimensionality, although it does not completely eliminate it.