Automated Machine Tool Prognostics for Turning Operation using Acoustic Emission and Learning Vector Quantization

Computer Numeric Control (CNC) machines are vital need of every manufacturing industry from small to large scale. Machine breakdown time or bad machining costs significantly with respect to time, quality and possible human labor injury that can be avoided by devising automated machine and tool condition monitoring systems to adapt automated preventive maintenance. This research is envisioned to device online tool condition monitoring system for CNC machine using airborne acoustic emission; a cheap yet highly accurate tool state prognostic scheme. Learning Vector Quantization (LVQ) neural network is trained on seven statistical features for two different types of machining materials (Aluminum and Mild Steel) with Carbide tool tip. Training accuracies of 95.33% and 94.33% are achieved with the proposed design for the two trails. The experimentation delivered promising industrial oriented performance scores with an efficient applicability.