Neural Network Based Force Modeling for Haptic Virtual Machining Simulation

In this paper, a novel haptic rendering method based on artificial neural network is proposed for turning simulation. In the proposed method, a two-layer neural network structure with three inputs and one output is used to model the cutting force. Levenberg-Marquardt algorithm is exploited as a learning strategy for training the neural network based on experimental machining data. Using the trained neural network, the relationship between simulated cutting force and cutting conditions can be obtained. Based on this method, a haptic virtual machining operation system is implemented and presented in this paper.