Abstract: The varying characteristic of machine tool chatter development is revealed, and a chatter monitoring parameter is presented. An adaptive computing ahead algorithm is developed, so as to predict the future values of the chatter monitoring parameter. The algorithm integrates a global generalization neural network with a locally tuning one. Using the proposed approach can exactly conduct the early forecasting of chatter.

Key words: Artificial neural network, Cutting dynamic stability, Early forecasting, Machine tool chatter