Title :
An adaptive nonlinear filter structure
Author :
Matthews, Michael B.
Author_Institution :
Swiss Federal Inst. of Technol., Zurich, Switzerland
Abstract :
A simple, nonlinear, state-space model based on saturating nonlinearities is presented and shown to be dense in the set of all fading-memory systems, i.e., it can uniformly approximate a fading-memory system over all time. The approximation can be made arbitrarily close by increasing the order of the model. Furthermore, the model has good stability and observability properties. The state and parameters of the approximation are estimated using the recursive prediction error algorithm
Keywords :
adaptive filters; filtering and prediction theory; observability; stability; state-space methods; adaptive nonlinear filter structure; approximation; fading-memory systems; observability; recursive prediction error algorithm; saturating nonlinearities; stability; state-space model; Filtering; Information processing; Nonlinear filters; Observability; Prediction algorithms; Recursive estimation; Signal processing; Stability; State estimation; State-space methods;
Conference_Titel :
Circuits and Systems, 1991., IEEE International Sympoisum on
Print_ISBN :
0-7803-0050-5
DOI :
10.1109/ISCAS.1991.176429
