Parameter Identification Method for a Periodic Time-Varying System Using a Block-Pulse Function

For periodic time-varying systems, a method of parameter identification based on the block-pulse function is presented. Firstly, the state-space equation of the system was expanded using the block-pulse function, then the recursion formula of the parameter identification of a time-varying system was obtained, according to the irrespective and orthogonal characteristics of the block-pulse function. This study provides a wide range of applications by saving time in calculation with a highly accurate method. The parameter identification was carried out by including the numerical simulation model of a three-degree freedom system and the vibration experiment results of an asymmetrical rotor system. The state space wavelet method and EMD method were compared cross-sectionally with the proposed method; this shows that the proposed method is accurate and effective, which makes it valuable in numerous applications. It also has a certain application value for several related projects.