Abstract—The objective of this study is to develop an effective approach for prediction of part dimensional accuracy and surface roughness in turning operations using neural network based multi-sensor integration strategy. The prediction system is built progressively by examining cutting parameters known to have influence on product quality from various aspects and making integration decisions step by step. The integration procedure begins by establishing the relationships between the cutting parameters and machined part quality and their sensitivity to the process conditions. Based on these results and using various statistical tools, variables selection, modeling and multi-sensor fusion procedures are executed. The results concerning finish turning of an Al alloy on a CNC turning machine demonstrate that the proposed approach can accurately predict on-line the part dimensional deviations and surface roughness under various machining conditions. The ANN based multi-sensor integration approach can be effectively and gainfully applied to in-process monitoring of product quality in turning operations because it includes the advantages of simple application, reduced modeling time, sufficient model accuracy and robustness.

Index Terms—cutting processes, part quality monitoring, quality prediction, sensor fusion, ANN

Cite: Abderrazak El Ouafi and Noureddine Barka, "An ANN Based Multi-Sensor Integration Approach for in-Process Monitoring of Product Quality in Turning Operations," Jounal of Automation and Control Engineering, Vol. 2, No. 3, pp. 289-293, September, 2014. doi: 10.12720/joace.2.3.289-293