Output characteristics modeling of fast tool servo based on neural network method

Flexure hinge mechanism driven by piezoelectric actuator is widely used in Fast Tool Servo (FTS) system. Most of the research focuses on constructing the model between the control voltage and output displacement. In this paper, the FTS is designed for compensating the machining error caused by flutter during turning. Therefore, the turning force should be considered as an additional load for the real time control system. This paper presents an output characteristics model of FTS based on the Neural Network model by analyzing the relationship among the output displacement, control voltage and external load. Finally, through the fitting plot and residual plot compared with the regression model, the accuracy and validity of the proposed method for the output characteristics model is demonstrated.