An Intelligent Grinding Wheel Wear Monitoring System Based on Acoustic Emission

Abstract

Acoustic emission (AE) signals can provide tool condition that is critical to effective process control. However, how to process the data and extract useful information are challenging tasks. This paper presented an intelligent grinding wheel wear monitoring system which was embedded in a surface grinding machine. An AE sensor was used to collect the grinding signals. The grinding wheel wear condition features were extracted by a proposed novel method based on statistics analysis of the average wavelet decomposition coefficient. The detailed signal characteristics during different wear condition are described. A BP neural network was used to classify the conditions of the grinding wheel wear. The inputs of the neural network were the three extracted features, and the outputs were three different states of grinding wheel condition, namely primary wear, intermediate wear and serious wear. The intelligent monitoring system was evaluated through grinding experiments. The results indicate that the effectiveness of the proposed method for extracting features of AE signals and developed intelligent grinding wheel wear monitoring system are satisfied.