Evolutionary modeling of ordinary differential equations for dynamic systems

摘要

This paper presents a new idea of modeling one-dimensional dynamic systems by higher-order ordinary differential equation (HODE) models in stead of by the ARMA models used in the traditional time series analysis. Accordingly, based on the idea of hybrid evolutionary modeling, we modify the HEMA algorithm in (Cao et al., 1998a; Cao et al., 1998b) to approach the modeling problem of HODEs for dynamic systems. Two applications of time series are used to demonstrate its effectiveness and advantages.