Quality characteristics prediction of complex mechatronics system based on symbolic dynamics and gray model

Abstract

This paper provides a symbolic quality characteristics prediction method of complex mechatronics system based on gray model. The time series data that are collected from the design, manufacture and service process are transformed to corresponding symbolic sequence data by using the symbolic modeling analysis method. The Shannon entropy that represents the variation of symbolic quality characteristics can then be calculated, and the histogram and main modes of symbolic sequences are obtained. A preliminary prediction of system quality characteristics is conducted based on the gotten main modes. By using the fitting capability of gray theory for the trend of data sequences, the GM(1,1) model is employed for the prediction modeling of symbolic quality sequences in this paper, and the prediction of corresponding quality characteristics of whole complex mechatronics system are implemented. Furthermore, the correlation of multiple quality characteristics is analyzed by relative entropy and gray correlation analysis method. The key quality fluctuation prediction of electric control and drive system of heavy machine tool by this method has shown that the presented method has high computational efficiency and large practical value.