Cellular Neural Networks for Pattern Recognition in Control Charts

The continuous detection and correction of unnatural process behaviours, due to special causes of variations, is a basic task in manufacturing systems to maintain any process stable and predictable. This difficult task is generally carried out by human experts. In this paper, a contribution is given to provide an intelligent tool to production analysts by synthesizing a Cellular Neural Network for the recognition of specific abnormal behaviours in control charts. The capabilities of the designed Cellular Neural Network are also investigated when detecting patterns corresponding to disturbed behaviours. Numerical results are reported to illustrate the performances of the designed network.