A Robust Methodology for Tool Condition Monitoring Using Spiking Neuron Networks

Artificial neural networks of sigmoidal and McCullochPitts neurons have found increasing favour in industry research because of their most attractive features, abstraction of hardly accessible knowledge and generalisation from distorted sensor signals. In recent years experimental evidence has been accumulating to suggest that biological neural networks, which communicate through spikes, use the timing of these spikes to encode and compute information in a more efficient way. In this paper it is presented a simplified version of a Self Organizing neural architecture based on Spiking Neurons and it is shown that this computational architectures have a greater potential to unveil embedded information in tool wear monitoring data sets and that smaller structures, compared to sigmoidal neural networks, are needed to capture and model the inherent complexity embedded in tool wear monitoring data. Additional, it is proposed a robust methodology based on tool wear estimation historical evolution that should improve estimation and predictive capabilities of Tool Condition Monitoring systems.