Constrained Kalman Filter in Structural Model Identification

In this paper, a constrained Kaiman filter is developed for the purpose of estimating the parameters of nonlinear structural hysteresis model. In this method, a constrained method is incorporated into the extended Kalman filter to develop convergence stability, precision and speed. At the same time, a global weighted iteration technique is used to tackle the problem of arbitrary initial values. It can be found that this method is highly efficient in identification of inelastic structures from both clean data and noise-corrupted data.