Data-model driven nonlinear dynamic modeling method for rolling linear guideway

The rolling linear guideway is a key component of the reciprocating linear motion system. Modeling the dynamic behavior of rolling linear guide is of great significance for analyzing the motion accuracy. The physical method was presented by Li et al. However, there is a obvious error between the theoretical solution and the experimental results. To reduce the error, the data-model driven modeling method is proposed for describing the nonlinear dynamic behavior of the rolling linear guideway in this work. The parameters in the physical model are identified by the intelligent optimization algorithm (in this work, Particle Swarm Optimization are employed) and experimental data. The objective function is to minimize the absolute error between the theoretical acceleration and the corresponding experimental one. The constraint condition is to meet the assembly requirements of the rolling linear guide. The experimental results shown that the solution accuracy of the proposed method is almost 100% higher than that of the theoretical solution method.