Nonlinear estimation of the Bouc-Wen model with parameter boundaries: Application to seismic isolators

The Bouc-Wen model is one of the most widely used parametric models of hysteresis. The exactness of the Bouc-Wen model depends on the correctness of the model parameters that are subjected to bound constraints in accordance with their physical meaning. Constraints and nonlinearities of the model introduce a complexity for the parameter estimation. In this paper, a constrained unscented Kalman filter (CUKF) is proposed in order to identify the hysteresis model parameters in a robust and reliable way. The adopted parameter estimator has been compared with the well-known extended Kalman filter (EKF) that is often used for nonlinear system identification. The effectiveness of the proposed approach has been verified by numerical simulations and experimental tests. In particular, for the experimental verification of the CUKF, the Bouc-Wen model has been adopted to describe the hysteretic shear behaviour of a seismic isolator. The results show that the CUKF provides better parameter identification than the EKF also taking into account parameter boundaries.