Seyed Taghi Akhavan Niaki - Department of Industrial Engineering, Sharif University of Technology
Masoumeh Nafar - Department of Industrial Engineering, Sharif University of Technology

In this paper, we consider monitoring correlated multi-attributes processes following multi-binomial distributions using artificial neural networks. In these processes, out-of-control observations are due to assignable causes coming from some shifts on the mean vector or covariance matrix of the proportion nonconforming of the attributes. We propose three neural networks. The first one detects whether the process is in control or not. If the process is out-of-control, the second and the third ones diagnose the process attribute(s) that has caused the out-of-control signal due to increase or decrease in proportion nonconforming of attributes respectively. In order to evaluate the performance of the proposed methodology we present two simulation studies.

Monitoring; Multi-attributes; Quality Control; Neural networks