The customer was determining the best way to thoroughly test a system using a very large number of potential input conditions and recorded response parameters. The tests were time consuming and done in large numbers so any reduction in the number of test input conditions or responses recorded would yield significant savings. All the inputs and responses were subject to constraints and weights.
SmartUQ’s optimized sampling technology was coupled with surrogate models in an iterative approach. The sampling tools were used to reduce the number of input conditions and measured responses and the emulation tools then predicted the rest.
The optimized iterative method was able to successfully reduce the total amount of sampling required while maintaining expected accuracy and confidence in testing results. Further, SmartUQ’s tools were much faster and with better scaling than the prior state of the art. These improvements allow SmartUQ to perform this kind of test optimization quickly and for much larger systems than were previously feasible.
