Application features of the nonlinear autoregression neural network at modeling of the machine tools thermal characteristics

Abstract

The research results of the nonlinear autoregressive neural network at predicting the machine tools thermal characteristics are presented. The temperature characteristic and a number of rotation frequencies, which implements the complex operating mode of the machine tool, are used as input signals. The output signals are represented by the temperature offsets sequence of the spindle head. An original approach to choosing a stable forecasting solution using a neural network is presented. Keywords nonlinear autoregressive neural network, thermal characteristics, machine tool