A Model-Based Iteration Method of the Estimation for the Bearings Dynamic Coefficients

Abstract

The journal bearings are widely used in the high load rotation machine. The dynamic coefficients of the bearings a critical factor in the fault diagnosis of the system. In this paper, a model-based iteration method is described to identify the bearing dynamic coefficients. The model system is created firstly by the finite element method. Secondly, the principle of the method is proposed combining with the Kalman filter and Tikhonov’s regularisation method, which uses the theoretical dynamic coefficients to estimate the new bearing displacement in the horizontal and vertical directions using single measurement and then recalculate the dynamic coefficients. Finally, to verify this method, numerical simulations are carried out under different rotation speeds. The results show that this method has high confidence intervals in the identification of both stiffness and damping coefficients. This method would provide a new route for the identification of the bearing dynamic coefficients and dynamic analysis for the rotor-bearing system.