Fuzzy Petri nets with neural networks to model products quality from a CNC-milling machining centre

Abstract

This paper presents a Petri net approach for the modeling of a CNC-milling machining centre. Next, by utilizing fuzzy logic with Petri nets (fuzzy Petri nets), a technique based on 9 fuzzy rules is developed. This paper demonstrates how fuzzy input variables, fuzzy marking, fuzzy firing sequences, and a global output variable should be defined for use with fuzzy Petri nets. The technique employs two fuzzy input variables (spindle speed and feed rate), throughout the milling operation in order to determine surface roughness. Additionally, a fuzzy Petri net is used with an artificial neural network for the modeling and control of surface roughness. Experimental results illustrate that the technique developed can be of benefit when the cutting tool has suffered damage throughout the milling operation. It also shows how the technique can react when the quality is high, medium, or low. The surface roughness represents the quality specification of products from the CNC-milling machining centre.