J.C. Martins (Portugal)
System identification. Kalman filters. Multiple model adaptive estimation. Parameter estimation.
The performance of the Multiple-Model Adaptive Es timation (MMAE) in the identification of the parame ters of a physical system using different noise scenarios is reported in this paper. The experiments target a second order mechanical system, representative of the physical systems appearing in many engineering situa tions. The MMAE algorithm performance is evaluated using the steady-state and time-varying form of the Kalman filters used by it. Then, a recursive algorithm for the re-discretization of the unknown parameters, in order to obtain an approximation of the true param eter when none of the models matches the system, is presented and its results reported for the steady-state case.
Important Links:
Go Back
