Learning input feature selection for sensor fusion in tool wear monitoring

We develop a neural network for on-line tool wear monitoring in metal cutting environments. After levels of tool wear are topologically ordered by the unsupervised Kohonen's Feature Map, input features from Acoustic Emission and force sensor signals are scaled by an additional supervised learning stage using Input Feature Scaling(IFS) algorithm developed in this work. In a machining experiment, without any off-line feature selection procedure, this neural network with the ability to learn feature selection achieves 94% and 92% accuracy for classification into two and three levels of tool wear, respectively. In conjunction with Kohonen's Feature Map, IFS is a practical and reliable pattern classifier for sensor fusion.