A practical Bayesian framework for backpropagation networks
Neural Computation
Bayesian Learning for Neural Networks
Bayesian Learning for Neural Networks
Hi-index | 0.00 |
Due to the rapid wear of the cutting tools when machining titanium alloy, tool condition monitoring (TCM) is most useful to avoid workpiece damage and maximise machining productivity. This paper uses sensor signals and feature analysis to identify a feature set for effective TCM. Firstly, basic requirements of sensor signals in tool condition identification are discussed, and the suitability of two candidate signals (acoustic emission and cutting force) commonly employed for machining monitoring are critically analysed. Their effectiveness in TCM is investigated based on extracted features of these signals, singly or in combination. Experimental results based on titanium machining, which is an expensive process with high tool wear, indicate that this proposed method is capable to determine a suitable sensing method and an effective feature set to identify tool condition.