A Non-Invasive Method to Evaluate Fuzzy Process Capability Indices via Coupled Applications of Artificial Neural Networks and the Placket–Burman DOE

The capability analysis of a process against requirements is often an instrument change. traditional and fuzzy approaches are the most useful statistical techniques for determining intrinsic spread controlled establishing realistic specifications use comparative processes. In industry, approach commonly used to assess impact continuous improvement projects. However, these methods evaluate indices could give misleading results because dataset employed corresponds final product/service measures. This paper reviews alternative procedure based on methodology involved in modeling design experiments. Firstly, model with reasonable accuracy developed using neural network approach. embedded graphic user interface (GUI). Using GUI, experimental carried out, first know membership function variability then include this model. Again, identifies improved operating conditions significative independent variables. A new generated conditions, including minimum error reached each variable. Finally, GUI get prediction response determined triangular predicted values. feasibility proposed method was validated random data set corresponding basis weight papermaking process. indicate that provides better overview performance, showing its true potential. can be considered non-invasive.