关键词: 刀具磨损, 神经网络, 实时性, 数控车床

Abstract: A method to estimate the tool wear in NC lathe, which based on multi-layered feed-forward neural networks with varying -factors BP learning algotithm extracts the information of tool wear from the changes of cutting force over the cutting process. The system consists of a dynamic and a static neural network. The dynamic one estimates the components of the tool wear and the static one update the dynamic neural network according to the changes of cutting force components. The adaptability of the dynamic neural network ensures that changes in machining parameters can be accommodated.

Key words: CNC lathe, Neural networks, Real-time, Tool wear