J.V. Abellan, F. Romero, H.R. Siller, C. Vila (Spain), and R. Morales-Menendez (Mexico)
Neural networks, decision-making algorithms, high perfor mance machining, cutting tool replacement, multi-sensor systems
A hybrid decision-support algorithm for cutting tool re placement in High-Performance Machining (HPM) is pre sented. The hybrid algorithm adjusts the reliability and failure functions of cutting tool wear and surface rough ness after each machining, using statistical data and data generated by an Artificial Neural Network (ANN). The ANN uses several inputs from a multi-sensor system. The ANN model predicts the probabilities of the cutting tool being worn and the surface roughness being outside spec ifications. The system validation showed that hybrid ap proaches for cutting tool replacement can lead to a higher utilisation of the cutting tool during HPM operations: up to 33.3% more time. Early results are promising.
Important Links:
Go Back
