On Optimal Estimation Problems for Nonlinear Systems and Their Approximate Solution
Authors: | Cervellera Cristiano, CNR National Research Council of Italy, Italy Alessandri Angelo, CNR National Research Council of Italy, Italy Grassia Aldo Filippo, CNR National Research Council of Italy, Italy Sanguineti Marcello, University of Genova, Italy |
---|
Topic: | 1.1 Modelling, Identification & Signal Processing |
---|
Session: | Nonlinear System Identification II |
---|
Keywords: | Optimal estimation, Lyapunov method, stability, neural networks, discretization |
---|
Abstract
An approach based on optimization is described to construct state estimators that provide a stable dynamics of the estimation error and minimize a L_p measure of the estimation error. The state estimator depends on an innovation function made up of two terms: a linear gain and a feedforward neural network. The gain and the weights of the neural network can be chosen in such way to ensure the convergence of the estimation error and minimize the L_p performance index, after a suitable discretization of the state and error space. Simulation results are reported.