Input estimation for nonlinear systems

Abstract

A novel input estimation inverse methodology of determining the time-varying exciting forces, named as the input, in a nonlinear system is presented. These forces are estimated from the measured dynamic response data of a nonlinear system using the approach. The algorithm includes the extended Kalman filter (KF) with a recursive estimator. The extended KF generates the residual innovation sequences. The estimator uses a least-squares algorithm, which employs the residual innovation sequences to evaluate the magnitudes and, therefore, the onset time histories of the exciting forces. Numerical simulations of a nonlinear system, an air-damped isolator model, demonstrate the accuracy of the proposed approach.