Abstract: High dimensional and complex nonlinear systems widely exist in practical engineering rotor systems such as aviation, aerospace and shipbuilding. In the key research fields such as aviation engine rotor systems and gas turbine rotor systems, these high-dimensional complex systems are often difficult to directly process data and analyze statistics. A method of construction of dynamical reduced order model of uncertain rotor systems and identification of model dispersion parameters was proposed to address the problem of high model dimensions in uncertain rotor systems. Firstly, based on the deterministic dynamic model and the static matrix reduction method, the deterministic dynamic reduction model was further improved. Then, based on random matrix theory and non-parametric dynamic modeling methods, an uncertain dynamic reduction model was proposed. Finally, the divergent parameters of the uncertain dynamic model were identified using the first-order critical speed, vibration mode, and experimental data of the system deterministic model. The identification results of divergence parameters were experimentally verified on a rotor experimental platform. The research results indicate that the difference between the experimental results and the mean vibration response after reducing the order is small, and the difference between the experimental results and the uncertain dynamic model is not more than 10%, indicating that the theoretical model used has high accuracy and reliability in describing the behavior of the rotor system. This model can provide a reference for further research on the uncertain rotor system of the model.
Key words: rotor-support system; uncertain rotor system; dynamical reduced order model; nonlinear system; dispersion parameter identification; nonparametric modeling method; matrix reduction method