A discussion of various approaches to the identification/approximation problem
Abstract
The problem of the identification/approximation of the transfer function of a system by its poles and residues when the noise-contaminated input and output are specified is examined. Particular attention is given to the Wiener filter theory, stochastic methods applied to system identification, and Prony's algorithm and its extensions. It is demonstrated that identical analytic equations can be obtained by different techniques even though they start with formulations based on different assumptions.
- Publication:
-
IEEE Transactions on Antennas and Propagation
- Pub Date:
- January 1982
- DOI:
- 10.1109/TAP.1982.1142738
- Bibcode:
- 1982ITAP...30...89S
- Keywords:
-
- Approximation;
- Linear Systems;
- Signal Processing;
- System Identification;
- Transfer Functions;
- Wiener Filtering;
- Data Sampling;
- Kalman Filters;
- Linear Prediction;
- Maximum Likelihood Estimates;
- Network Synthesis;
- Stochastic Processes;
- Time Series Analysis;
- Transient Response;
- Communications and Radar