- 446 Downloads
A few mostly numerical examples show how to deal with challenging situations. If a system is affected by several different factors, such as components, subsystems, combinations, or options, an orthogonal array is applied to use minimal data as a generally rule. With the ongoing development aiming at higher system complexity, both, soft- and hardware fall within this category. It is not practicable to test all combinations. Assigning on-off situations as factors to an orthogonal array of adequate size is considerably easier and faster with unambiguous results. If a decision needs to be made it suffices to choose the result with the higher SNR index. For investigation and analysis, a system can be stressed to its limits if the model of prediction is confirmed. There is a chance to unravel basic technical, physical, or chemical relations to substitute empirical constants.