Surveillance intelligente des outils de coupe

Abstract

Cutting tool monitoring has become an important part of automation in manufacturing industries, especially where product quality and efficiency are in high demand. Conventional cutting tool monitoring methods use imaging and machining parameters such as cutting force, torque, feedrate, etc. to detect the presence of wear in the tools. A cutting tool condition monitoring system is vital to the machining process on CNC machines since tool degradation directly affects the integrity and cost of manufactured products. Therefore, an intelligent monitoring strategy is indispensable to improve the availability of the machining system, reduce downtime costs and improve operational reliability. We propose in this study to use various strategies for monitoring the state of cutting tools through the exploitation of vibration signals recorded during machining. Several temporal, frequency, or even time-frequency indicators will be used during the different phases of feature extraction, dimensionality reduction and classification. The final objective is to find a methodology, based on vibrations, precise and effective in terms of time for the monitoring of the cutting tools of a numerically controlled (CNC) machine.