Data-Driven Parameters Identification Method for Freeplay Nonlinear System

In this paper, a data-driven method is proposed for the identification of freeplay nonlinear structure parameters. This method combines two paths, including theoretical derivation and data-driven. The function relation between the equivalent stiffness and the freeplay nonlinear parameters is derived based on the harmonic balance method. Then the nonlinear parameters such as frequency response function, natural frequency and equivalent stiffness of the dynamic system are obtained gradually in the frequency domain. Finally, the freeplay nonlinear structure parameters are identified by Particle Swarm Optimization method. By comparing the theoretical values with the identified values of equivalent stiffness and freeplay nonlinear structure parameters, the effectiveness and accuracy of the identification method are verified.