Genetic algorithm for dynamic parameters estimation of the machine tool worktable using the residual vibration signal

Abstract

In this article, the residual vibration signal of the worktable of the machine tool is used to identify its dynamic parameters. By analyzing the time-domain characteristics of the residual vibration signal, it is found that the signal has the similar attenuation characteristics to the impulse response of the second-order underdamped system, so a new method for identifying the dynamic parameters of the worktable is proposed. In this method, the impulse response model of the second-order underdamped system is used to describe the residual vibration signal, so the parameter identification problem based on the time-domain signal is transformed into the optimization problem of the model parameters. Then, the dynamic parameters can be determined by using the genetic algorithm. The results of the genetic algorithm are compared with the results of the autoregressive moving average method, which is a commonly used parameter identification method based on the time-domain signal. It was found that the results of the model-based method proposed in this article are more accurate than the autoregressive moving average method.