Supervision controller for real-time surface quality assurance in CNC machining using artificial intelligence. (January 2019)

• Record Type: Journal Article
• Authors: Moreira, L.C.; Li, W.D.; Lu, X.; Fitzpatrick, M.E.
• Is Part Of: Computers & industrial engineering. Volume 127(2019)
• Page Start: 158
• Page End: 168
• Publication Date: 2019-01
• Subjects: CNC machining -- Surface roughness -- Quality control -- Smart manufacturing; Engineering -- Data processing -- Periodicals; Industrial engineering -- Periodicals; 620.00285
• DOI: 10.1016/j.cie.2018.12.016
• Languages: English
• ISSNs: 0360-8352

Abstract

Highlights: Innovative supervision controller for real-time adjustment on machining parameters. The controller is innovative in closed-loop control to achieve surface quality. Results demonstrate the approach is effective for high-quality machining processes. Abstract: A major challenge for Computer Numerical Control (CNC) machining is how to manufacture high-quality workpieces effectively. Consequences of poor surface quality incur re-processing and higher wastes generating negative impacts on production costs and profitability. The complex relationships between surface quality and machining parameters could overwhelm machinists' capabilities to correctly select machining parameters to produce satisfied quality of machined workpieces. This paper presents a novel approach of designing an intelligent supervision controller for real-time adjustments on feed rate and spindle speed to achieve desired surface quality of machined workpieces. The controller is an innovative model-based closed-loop system, consisting of a surface roughness prediction model and a multi-variable controller, to ensure real-time improvements on surface quality during machining processes. A case study based on milling processes for BS EN24T steel alloy has been used for testing and validating the approach. Simulation results show that the controller significantly reduced the difference between required and predicted surface roughness from 3.6 μm (based on planned parameters) to 0.12 μm (after the supervision controller adjustments). The results demonstrate that the proposed approach can effectively support high-quality machining processes.